UN Tourism | Bringing the world closer

The first global dashboard for tourism insights.

  • UN Tourism Tourism Dashboard
  • Language Services
  • Publications

share this content

  • Share this article on facebook
  • Share this article on twitter
  • Share this article on linkedin

UN Tourism Data Dashboard

The UN Tourism Data Dashboard – provides statistics and insights on key indicators for inbound and outbound tourism at the global, regional and national levels. Data covers tourist arrivals, tourism share of exports and contribution to GDP, source markets, seasonality and accommodation (data on number of rooms, guest and nights)

Two special modules present data on the impact of COVID 19 on tourism as well as a Policy Tracker on Measures to Support Tourism

The UNWTO/IATA Travel Tracker

The UN Tourism/IATA Destination Tracker

UNWTO Tourism Recovery Tracker

UN Tourism Recovery Tracker

International Tourism Results

  • International tourist arrivals and receipts and export revenues
  • International tourism expenditure and departures
  • Seasonality
  • Tourism Flows
  • Accommodation
  • Tourism GDP and Employment
  • Domestic Tourism

International Tourism and COVID-19"

International Tourism and COVID-19

  • The pandemic generated a loss of 2.6 billion international arrivals in 2020, 2021 and 2022 combined
  • Export revenues from international tourism dropped 62% in 2020 and 59% in 2021, versus 2019 (real terms) and then rebounded in 2022, remaining 34% below pre-pandemic levels.
  • The total loss in export revenues from tourism amounts to USD 2.6 trillion for that three-year period.
  • International tourist arrivals reached 88% of pre-pandemic levels in January-December 2023

COVID-19: Measures to Support Travel and Tourism

travel social data

No spam. We promise.

Login to download report

travel social data

World Travel & Tourism Council Research Hub

Welcome to WTTC's global Research Hub. Here you will find all our reports which provide critical evidence to empower the public and private sectors, and help to shape policies and investments that bolster Travel & Tourism.  We analyse Travel & Tourism's economic and employment impact in 185 economies, 82 cities and 26 regions worldwide. We also produce reports on the environmental and social impacts of Travel & Tourism, as well as thought leadership reports on themes such as diversity, equity, inclusion and belonging, and retail tourism. Explore full reports (free for WTTC members and purchasable for non-members) or discover valuable insights in our library of freely available economic factsheets.

Featured Reports

travel social data

Nature Positive Travel & Tourism in Action

travel social data

Responsible Artificial Intelligence (AI) : Overview of AI Risks, Safety & Governance

Click on the below links to select a report or factsheet for a particular region

(By region or economy)

(By region, economy or city)

Nature Positive Travel & Tourism in Action

Artificial Intelligence (AI): Global Strategies, Policies & Regulations

Responsible Artificial Intelligence (AI) : Overview of AI Risks, Safety & Governance

Navigating the Sustainability Journey: The Impact of Mandatory Reporting on Travel & Tourism

Introduction to Artificial Intelligence (AI) Technology: Guide for Travel & Tourism Leaders

Introduction to Artificial Intelligence (AI) Technology: Guide for Travel & Tourism Leaders

Artificial Intelligence (AI) in Action: Use Cases & Impacts of AI in Society, Business and Travel & Tourism

Artificial Intelligence (AI) in Action: Use Cases & Impacts of AI in Society, Business and Travel & Tourism

 The Social Impact of Global Tourism

The Social Impact of Global Tourism

Sustainability Leadership Case Studies / Iberostar: Roadmap for Decarbonisation

Sustainability Leadership Case Studies / Iberostar: Roadmap for Decarbonisation

Water Roadmap for Travel & Tourism

Water Roadmap for Travel & Tourism

Creating Belonging: Diversity, Equity, Inclusion, & Belonging in Travel & Tourism

Creating Belonging: Diversity, Equity, Inclusion, & Belonging in Travel & Tourism

Unlocking Opportunities for Travel & Tourism Growth in Africa

Unlocking Opportunities for Travel & Tourism Growth in Africa

The Environmental Impact of Global Tourism Report

The Environmental Impact of Global Tourism Report

Travel & Tourism Environmental & Social Impact Factsheets

Travel & Tourism Environmental & Social Impact Factsheets

Sustainability in Numbers: Travel & Tourism in Puerto Rico

Sustainability in Numbers: Travel & Tourism in Puerto Rico

Global Retail Tourism Trends and Insights: Executive Summary

Global Retail Tourism Trends and Insights: Executive Summary

Global Retail Tourism Trends and Insights: Full Report

Global Retail Tourism Trends and Insights: Full Report

Economic Impact Report 2023 Global Trends

Economic Impact Report 2023 Global Trends

Cities Economic Impact 2022: Executive Summary

Cities Economic Impact 2022: Executive Summary

Cities Economic Impact Report 2022

Cities Economic Impact Report 2022

Nature Positive Travel & Tourism: Travelling in Harmony with Nature

Nature Positive Travel & Tourism: Travelling in Harmony with Nature

Toolbox of Nature Positive Tourism Resources

Toolbox of Nature Positive Tourism Resources

Enhancing Resilience: To drive sustainability in destinations

Enhancing Resilience: To drive sustainability in destinations

Destinations 2030: Global Cities’ Readiness For Sustainable Tourism Growth

Destinations 2030: Global Cities’ Readiness For Sustainable Tourism Growth

A Net Zero Roadmap for Travel and Tourism: Proposing a New Target Framework for the Travel & Tourism Sector

A Net Zero Roadmap for Travel and Tourism: Proposing a New Target Framework for the Travel & Tourism Sector

travel social data

Have a question, technical issue or can't find a report?

Travel & tourism global footprint interactive website.

travel social data

Influencers and Social Commerce in Travel

A preview of one of phocuswright's travel innovation and technology trends 2023 : the future of social media, influencers and social commerce in travel.

By Robert Cole

travel social data

The goal of travel industry marketers has always been to build efficient direct-to-consumer relationships by cost-effectively expanding brand reach, enhancing traveler engagement and maximizing customer lifetime value. Data privacy regulation (e.g., the impending demise of third-party cookies) is disrupting digital advertising by impairing the relevance of important targeting data, which will further elevate the importance of direct consumer relationships. 

Travel brands, particularly hoteliers, compete with online travel agencies (OTAs) that transact approximately 50% of U.S. online hotel bookings . Accommodation bookings also generate the lion’s share of OTA profitability. In the highly fragmented tours and activities space, 52% of bookings are through intermediary channels with only about 25% of bookings currently processed online. 

The confluence of these forces creates a catalyst to engage directly with consumers through social commerce – a rapidly growing sector where travel currently lags. According to one projection, U.S. social commerce will more than double from $37 billion in 2021 to $80 billion in 2025 , growing its share of total e-commerce sales to 5.2%.

This preview looks at how social marketing has evolved, how consumers interact with brands and nuanced differences for travel. Get the full report for a deep analysis of:

  • How loyalty and marketing has evolved
  • How influencers are compensated and create revenue streams
  • Lessons from fashion and beauty
  • Nuanced strategies of different travel brands
  • A breakdown of lodging affiliate programs
  • Aligning attribution with conversion
  • Capturing data while respecting the privacy of customers, influencers and brands
  • Successful travel brand advocacy platforms and campaigns
  • And much more

Get the full report The Future of Social Media, Influencers and Social Commerce in Travel here .

travel social data

Social Commerce Goes Beyond Social Media Marketing 

Where traditional social media campaigns may feature links to a new product, a targeted persona landing page, a seasonal sale or special coupon code, social commerce campaigns promote specific products through deep links into specific product pages. They often offer inventory and pricing, with “buy now” as the call to action. 

Influencer-led communities epitomize democratized media. Creators are self-contained lifestyle brands primarily interested in supporting products and experiences that closely align with their lifestyles and the sensibilities of their followers. Product advocacy extends beyond the influencers through their audience.

travel social data

Image captions

travel social data

Nuanced Differences for Travel

Followers thrive on relevance and authenticity. Just like fit is important in fashion, the “fit” of a destination, hotel or activity is just as vital to a follower when searching for trip inspiration. 

The equivalent of the try-on for travel is the site visit. Photos and videos taken on-site deliver far better engagement when compared to textual mentions associated with stock or brand-provided photos. This need for authentic social proof raises the bar considerably for travel brands when selecting influencers and planning campaigns. 

travel social data

Keeping Receipts 

Social commerce connects the dots between top-of-funnel inspiration and product discovery campaigns to mid-funnel merchandising, referral and retargeting strategies, and ultimately to bottom-funnel discount and shopping cart conversion tactics. 

travel social data

Big Returns From Average Visitors 

Alternate approaches can leverage social commerce by engaging regular travelers. Brand-backed content marketing efforts lack authenticity, as travelers suspect products are always portrayed under ideal, staged conditions. Instead, consumer-created content is not only distributed through the traveler’s social network, but the best submissions may be curated by the brand for use in future marketing campaigns. 

The most effective marketing strategies are informed by performance metrics that help creative teams identify new opportunities to engage consumers. 

travel social data

Social Commerce and the Future of Travel

As media, finance and identity continue to evolve, the ability to accurately attribute lead sources and conversion associated with influencers, channels and promotional campaigns will be essential to optimize marketing campaign strategies and budget efficiency across every component of a travel itinerary. 

travel social data

Influencer-driven social commerce represents a major opportunity for a travel industry striving to crack the code of developing direct distribution channels, nurture loyalty and improve marketing efficiency. New social platforms may rise and fall, but the key to success will be deploying broadly applicable, yet customizable marketing infrastructure that bridges campaign, booking, operations, loyalty and analytics. 

The full report contains deep analysis of:

Each year, Phocuswright's expert analysts identify the technology and innovation trends that will influence travel significantly in the coming year and beyond. This year, we’re exploring the growing roles of social media and Web3 in travel, addressing the realities of sustainability and our fragmented technology landscape, and pondering the impact of the next game-changing technologies like generative AI and eVTOLs. 

This overview article features brief introductions to the eight trends that we will cover in the coming months.

  • The Future of Social Media, Influencers and Social Commerce in Travel 
  • Web3 Is Proving Itself in Travel 
  • Green Travel Innovation Now (Yes, Now!) 
  • No Travel Experience Necessary: More Outsiders Enter the OTA Market
  • Generative AI: Transforming the Travel Cycle
  • Real-time Revolution in Hotel Operations 
  • eVTOLs in Travel: Viable Addition or Flights of Fancy? 
  •  Super Apps’ Secret Sauce 

Get the full overview article here .

Watch the online event that covered each trend, presented by Phocuswright analysts:

travel social data

WEEKLY RESEARCH INSIGHTS

We dig deep to give you the data and trends that drive the travel, tourism and hospitality industry.

Get the latest in travel industry highlights with our free weekly research articles and more. Sign up to get the latest delivered directly to your inbox.

FOR MORE INSIGHTS

See all of Phocuswright's free research insights here .

Sign up to get the latest delivered directly to your inbox.

Open Access Research Subscription

Research is our priority. Our Open Access research subscription puts the world’s most comprehensive library of travel research and data visualization at your fingertips.

Clients have relied on Phocuswright's deep industry knowledge for over 25 years to power great decisions, help justify a pitch, build a strategic plan and elevate any presentation through trusted research and data. When companies and executives reference Phocuswright, they gain the trust of an industry keen on data, trends and analytics.

See the full benefits of an Open Access subscription here .

Feasibility of estimating travel demand using geolocations of social media data

  • Open access
  • Published: 26 January 2021
  • Volume 49 , pages 137–161, ( 2022 )

Cite this article

You have full access to this open access article

travel social data

  • Yuan Liao   ORCID: orcid.org/0000-0002-6982-1654 1 ,
  • Sonia Yeh 1 &
  • Jorge Gil 2  

4520 Accesses

13 Citations

11 Altmetric

Explore all metrics

Travel demand estimation, as represented by an origin–destination (OD) matrix, is essential for urban planning and management. Compared to data typically used in travel demand estimation, the key strengths of social media data are that they are low-cost, abundant, available in real-time, and free of geographical partition. However, the data also have significant limitations: population and behavioural biases, and lack of important information such as trip purpose and social demographics. This study systematically explores the feasibility of using geolocations of Twitter data for travel demand estimation by examining the effects of data sparsity, spatial scale, sampling methods, and sample size. We show that Twitter data are suitable for modelling the overall travel demand for an average weekday but not for commuting travel demand, due to the low reliability of identifying home and workplace. Collecting more detailed, long-term individual data from user timelines for a small number of individuals produces more accurate results than short-term data for a much larger population within a region. We developed a novel approach using geotagged tweets as attraction generators as opposed to the commonly adopted trip generators. This significantly increases usable data, resulting in better representation of travel demand. This study demonstrates that Twitter can be a viable option for estimating travel demand, though careful consideration must be given to sampling method, estimation model, and sample size.

Similar content being viewed by others

travel social data

Smart Transportation Systems in Smart Cities: Practices, Challenges, and Opportunities for Saudi Cities

travel social data

The accuracy of crime statistics: assessing the impact of police data bias on geographic crime analysis

travel social data

Can passenger flow distribution be estimated solely based on network properties in public transport systems?

Avoid common mistakes on your manuscript.

Introduction

Travel demand estimation is essential for urban planning and management of transportation networks. The time series of visits to various locations by individuals are aggregated to study the flows of people between different zones/regions. Based on the spatio-temporal scale of the aggregation, an origin–destination (OD) matrix can be constructed with the origins and destinations of all trips. These OD matrices are particularly important for representing travel demand (Calabrese et al. 2011 ). Traditionally, the estimation of OD matrices relies on input data from household travel surveys, censuses, and traffic surveys that feature representative populations and detailed information about travel mode and trip purposes. However, data collection frequency, methods, and data availability vary across countries (and across cities within a country), making it difficult to interpret the results. For example, in the UK and the Netherlands the travel surveys are done annually, but that is an exception. Other places do not do them regularly, if at all. Portugal had one travel survey carried out for two metro areas in 2017, but nothing more since. Otherwise, mobility is derived from the census data (carried out every 10 years), but that offers a different resolution since it is not based on travel diaries. On top of these issues, the costs of these surveys are increasing, while the response rates are decreasing over time (Yue et al. 2014 ), making it hard to keep the travel demand models up to date. Emerging data sources associated with mobile/smart phones are increasingly leveraged to overcome these drawbacks.

In the last decade, the emerging data sources have significantly improved our understanding of travel behaviour (Gonzalez et al. 2008 ; Song et al. 2010 ; Barbosa et al. 2018 ) and have brought new opportunities for travel demand modelling (Anda et al. 2017 ). Common emerging data sources are call detail records (CDR) (Calabrese et al. 2011 ), smart card data, GPS-enabled devices, and geotagged social media, e.g., Twitter (Lee et al. 2019 ; Hasnat and Hasan 2018 ).

Alongside the development of information and communication technologies (ICT), interest in online social media services, e.g. Twitter, has grown among the transportation research community (Rashidi et al. 2017 ). A tweet typically contains multiple components that can be useful for transport research, including text, hashtag, location, and timestamp. When users choose to have their location reported when sending out tweets, these are called geotagged tweets . Despite geotagged tweets accounting for a small proportion (1–3%) of all tweets (Morstatter et al. 2013 ), these check-ins provide precise location information and have increasingly been used for estimating mobility and travel demand either at the global (e.g. Hawelka et al. 2014 ) or regional level (e.g. Yang et al. 2015 ).

In the estimation of travel demand, two forms of data are often used: longitudinal and lateral. A longitudinal data set is characterised by long-term (more than 24 h) and continuous observations focusing on a group of participants. A lateral data set is often collected based on a particular area, such as a city or a country, during a short to medium time period, and it usually covers a larger population. Thus, the data offer either broader or longer coverage, but rarely both.

Geotagged tweets can be obtained in three ways: (1) Purchase the complete set of public tweets from Twitter Firehose (Twitter 2019c ); (2) Access the Streaming API to get a maximum of 1% of the public tweets (Twitter 2019a) ; (3) Access the user timeline by user name/ID to get a maximum of 3200 historical tweets that are set by the user as publicly accessible (Twitter 2019b ). Different collection channels of geotagged tweets correspond to different data forms. Sampling methods (1) and (2) collect geotagged tweets generated within a specified region, while sampling method (3) collects data from user timelines without any spatial boundaries.

Geotagged tweets collected from Twitter Firehose and Streaming API are often limited to a geographical bounding box yielding a lateral data set. It covers a large number of Twitter users but takes time to accumulate enough samples for each individual, and movements outside or across the bounding box are not captured (Liao et al. 2019 ). Alternatively, by accessing User Timeline API, all publicly available historical tweets by a specific user can be collected to form a longitudinal record of individual trajectories without any geographical boundaries. Longitudinal geotagged tweets are collected without being constrained to a specific area, but typically with a smaller number of individuals, albeit a much larger overall sample size (one to two orders of magnitude more samples per user).

Most studies use geotagged tweets in the lateral form, focusing on a specified area in line with the spatial scale of policy-making and urban planning. For example, one study modifies a classic movement model by integrating locations posted on Foursquare (which Twitter integrates) for origin–destination estimation in Austin, Texas (Jin et al. 2014 ). Longitudinal data can also be scaled up to large numbers of Twitter users to study the OD flows between global cities (Lenormand et al. 2015 ).

One recent literature review shows that experts are optimistic about the usefulness of such data sources for modelling travel behaviour (Rashidi et al. 2017 ). Compared with the other data sources, geotagged tweets have several strengths: long collection duration, large number of studied individuals, large spatial coverage, ease of access, low cost, and accurate location information. The low cost of retrieving geotagged tweets makes them especially appealing compared to other data sources (Rashidi et al. 2017 ). The data source is free to access, and it provides precise location information with a spatial resolution of around 10 m compared with 100–200 m for call detail records (CDR) (Jurdak et al. 2015 ). Moreover, it is relatively scale free, i.e. analyses can be done with any desired time frame and spatial boundaries based on the research question at hand (Liao and Yeh 2018 ).

Despite the wide applications, rigorous cross-validation of the use of emerging data sources, such as geotagged social media data, to approximate the travel demand, and their robustness across spatial and temporal scales is still lacking. The main criticism of Twitter data pertains to two aspects: a biased population representation, and low and irregular sampling. Geotagged tweets can capture movements over multiple years and include overseas visits, but the data are “sparse”, thus the picture of actual movements is incomplete (Liao et al. 2019 ). There have been studies comparing multiple data sources to identify/adjust the biases (e.g. Wesolowski et al. 2013 ; Tasse et al. 2017 ) and to validate against “ground truth” (e.g. Lee et al. 2019 ). It is worth noting, however, that the “ground truth” is also an incomplete picture of reality, as it is, at best, based on the knowledge from well-recognised but limited data collection and established modelling techniques.

This study attempts to comprehensively examine the validity of using geotagged Twitter data for travel demand estimation by comparing Twitter data sets with established data sources. We first compare the empirical trip records with respect to the commuting travel demand and the overall travel demand for an average weekday. We then create gravity models based on Twitter data to estimate the overall travel demand at both the national (long-distance travel above 100 km) and city level. Finally, we compare Twitter-based OD matrices and trip distance distributions with those from the other established sources using spatially weighted structural similarity index and Kullback–Leibler divergence, respectively.

The main contributions of this study lie in the quantification of the feasibility of using geolocations of Twitter data for estimating commuting demand and the overall travel demand, given different sample sizes, sampling methods of Twitter data, and spatial scales. In addition, we develop a novel approach using geotagged tweets as attraction generators as opposed to the commonly adopted trip generators. This significantly increases usable data, resulting in better representation of travel demand and the promise for using Twitter data at a finer spatiotemporal resolution.

The remainder of this paper is organised as follows. “ Related work ” section reviews work related to travel demand estimates using social media data and outlines the objectives of the present study. “ Data description ” section describes the data, and “ Methodology ” section describes the methods used. The results are presented in “ Results ” section, and  “ Discussion ” section discusses the findings. “ Conclusion ” section  concludes and identifies future research needs.

Related work

Modelling travel demand.

For travel demand estimation, one needs to first extract activities and trips where Twitter data have proven useful for both conventional four-step modelling and activity-based modelling by providing inferred activities and trips. There has been increased interest in developing methods to infer this information using social media check-in data, such as Twitter data. One recent study has demonstrated that Twitter data can be integrated with an household travel survey to improve the quality of OD matrices (Cheng et al. 2020 ). Constructing activity-based models requires trip purpose, departure time, and socioeconomic attributes of travellers, among other attributes. The content of geotagged tweets is often used with text mining to extract those attributes, e.g., the activity purposes such as work and leisure and the socio-economic profile of Twitter users (Hasan and Ukkusuri 2014 ; Abbasi et al. 2015 ; Maghrebi et al. 2015 ).

The methodology of four-step travel demand modelling (McNally 2007 ) consists of trip generation and trip distribution as the first two steps. It starts from the definition of a trip , which is the connection between two consecutive stays generated by the same individual. This individual refers to a phone user when using CDR data (Calabrese et al. 2011 ), or a survey participant from a one-day travel diary. When it comes to geotagged social media data, a trip is generally defined in the literature as the connection between two consecutive geotagged tweets generated by the same Twitter user. However, due to the sparsity and incomplete trajectory of geotagged tweets, the time interval between two consecutive geotagged tweets can be extremely long (from a few hours to several weeks/months), while the air distance can be close to zero. Therefore, in this context, “ displacement ” is a more appropriate term than the traditional sense of the trip. Despite a displacement in geotagged tweets being different from a record in a travel diary, existing literature often uses these two terms interchangeably.

Trip generation

Trip generation involves the estimation of the number of trips produced by and attracted to each zone, either using empirical data directly, or modelled results based on zonal demographics and land use information.

Social media data such as displacements in Twitter data need to be processed to become trips. Gao et al. ( 2014 ); Kheiri et al. ( 2015 ) and Lee et al. ( 2019 ) propose displacement conversion where they filter out those displacements with time intervals longer than a selected time threshold, e.g. 4 h, 12 h, or 24 h. However, this time threshold is arbitrary and the choice results in a massive reduction of available data.

Instead of geotagged displacements, one can model destination choices to estimate zonal attractiveness. Hasnat et al. ( 2019 ) applied Twitter data together with census tract data for modelling travellers’ destination choice behaviour, which suggests that Twitter data can be utilised effectively for modelling destination choices that reflect the attractions of zones. Molloy and Moeckel ( 2017 ) develop a long-distance destination choice model using Foursquare check-ins whose results suggest that check-ins from social media platforms can improve destination choice models, particularly for leisure travel.

Trip distribution

Trips are further aggregated to OD zones depending on the spatial scale. The step of trip distribution assigns trips produced by each zone to each of the other zones where these trips are attracted to Anda et al. ( 2017 ). There are many models to assign the number of trips between each pair of OD zones. In a study by Yang et al. ( 2015 ) of the Chicago metropolitan region, daily check-ins from Foursquare are used to estimate the productions and attractions in each traffic analysis zone as inputs to gravity models for estimating trip distribution. By further calibrating against the OD matrix from other data sources such as CDRs, they demonstrate how to use gravity models with check-in data to estimate the OD matrix. Kheiri et al. ( 2015 ) use the radiation model, rank-based model, and population-weighted opportunities model to distribute the trips generated with Foursquare check-ins to estimate the OD matrix.

Commuting travel demand estimation

Estimating the OD matrix according to trip purpose points toward more specific applications. Commuting flows account for a large share of total trips, therefore they attract more attention. For example, Zagatti et al. ( 2018 ) use CDRs to estimate an OD matrix of commuting flows. For social media data, some data sources have trip purposes (activity types), such as Foursquare, while Twitter data do not directly provide this information. With a small share of check-ins at home/workplace from Foursquare when compared with the actual daily mobility, Yang et al. ( 2015 ) focus on non-commuting trips. To construct OD matrices of commuting flows with geotagged tweets or CDRs, one needs to detect home/workplace when the trip purpose is not explicitly given. Schneider et al. ( 2013 ) assume that the most visited location during weekends and 7 pm–8 am on weekdays is the home location and the second most visited location during 8 am–8 pm on weekdays is identified as one’s workplace. Combining such temporal rules and visiting frequency, this method has been widely used to identify the home/workplace through social media data (Wang et al. 2018 ; Osorio-Arjona and García-Palomares 2019 ), sometimes together with land-use information (Osorio-Arjona and García-Palomares 2019 ).

Efforts that infer the home/workplace from geotagged tweets must consider the behavioural bias of people geotagging consciously and intentionally in uncommon places to communicate and show where they have been (Tasse et al. 2017 ). Home and workplace are at the opposite extreme, i.e., they are the most common places that people visit on a daily basis. A preliminary comparison between Twitter data and the national travel survey suggests that the low probability of reporting home and workplace implies that further scrutiny of the validity of estimating commuting-OD matrices based on geotagged tweets is required.

Validation against other data sources

Researchers have devoted efforts to validating geotagged tweets with other data sources. A study focusing on the U.S. found that densely populated regions and males were over-represented among Twitter users (Mislove et al. 2011 ). In addition, there are two possible types of behavioural distortion for Twitter users who geotag: only tweeting at specified locations or times, and geotagging only certain or all of the tweets.

When cross-validating against data with higher temporal resolution such as CDR (Lenormand et al. 2014 ), good agreement is generally found regarding, for instance, trip distance distribution. When validating geotagged tweets against travel surveys, studies show that geotagged social media data capture the displacement distribution, length, duration, and start time of trips reasonably well for the purpose of inferring individual travel behaviour (Zhang et al. 2017 ; Liao et al. 2019 ). Validations using CDR need careful interpretation, as CDR and geotagged tweets are both passive data collection methods that share some similar shortcomings.

Good agreement on fundamental indicators of individual travel behaviour does not necessarily guarantee a good proxy for the travel demand at the population level. Some studies comparing geotagged tweets with traffic data (Ribeiro et al. 2014 ) and travel-demand data (Lee et al. 2015 , 2019 ; Yang et al. 2015 ) have generally achieved good results. However, as pointed out recently by Lee et al. ( 2019 ), the sparsity of geotagged tweets leads to sparse OD matrices and therefore cannot replace other travel demand forecasting methods for state-wide travel models.

Study objectives

The work comparing geotagged tweets with other data sources for travel demand estimation still lacks systematic rigour in at least four areas: (1) Commuting travel demand. The basic temporal technique to identify home or workplace has been widely applied for deriving commuting trips. Our preliminary results from previous analyses suggest that identifying home and workplace locations through geotagged tweets gives mixed results and the reliability of the method requires further scrutiny; (2) Spatial scale . Most studies look at pre-selected regions without exploring the effects of spatial scales on travel demand estimation, whereas we hypothesise that the feasibility of using Twitter data for travel demand estimation can depend on the scale; (3) Sampling methods . The existing literature is not clear on how different sampling methods (region-based vs. user-based) affect the validity of using geotagged tweets to estimate travel demand; (4) Sample size . It remains unclear how the sparsity of Twitter data affects the validity of using it for travel demand estimation.

To fill these gaps in the literature, we systematically examine the validity of using geotagged tweets collected within a specified region, and from user timelines, to approximate the OD matrix at different spatial scales. We compare these Twitter-based OD matrices with the Swedish national travel survey and output from Swedish Transport Administration (Trafikverket) traffic models. Specifically, we attempt to answer the following questions:

Are Twitter data a feasible source for representing commuting travel demand?

Can geolocations of Twitter data be used to create models for travel demand estimation?

How do spatial scale, sampling method, and sample size of Twitter data affect its representativeness for travel demand?

Data description

This study focuses on Sweden as a whole and on Greater Gothenburg, located in western Sweden. Sweden is a European country with a population of 10.2 million in 2019 and the GDP per capita was 54.6 kUSD in 2018 (Statistics Sweden). Gothenburg is its second largest city for which Greater Gothenburg covers its metropolitan area with a population of around 1 million.

Specifically, four datasets have been used in this study. Two Twitter datasets collected using different sampling methods: lateral geotagged tweets (Twitter LT), and longitudinal geotagged tweets (Twitter LD). And two datasets with which the Twitter data are compared: the Swedish National Travel Survey; and OD matrices from the Sampers model, a traffic simulation model with the travel demand module embedded, developed by the Swedish Transport Administration. The traffic zones used by Sampers are illustrated in Fig. 1 for two spatial scales: Greater Gothenburg (city level) and Sweden (national level). Detailed descriptions of each dataset are presented in this section.

figure 1

Geotagged tweets in traffic zones from Twitter LD (blue) and LT (green) for Sweden (left) and Greater Gothenburg (right). (Color figure online)

Twitter data

Lateral geotagged tweets (twitter lt).

We purchased data from Gnip, a Twitter subsidiary, during a 6-month period (20 December 2015–20 June 2016) within the geographical bounding box of Sweden (Jeuken 2017 ; Liao et al. 2019 ). Gnip sells complete historical tweets in bulk and provides access to the Firehose API.

Longitudinal geotagged tweets (Twitter LD)

We identify 7773 top geotag users from Twitter LT who geotagged their tweets most frequently during that 6-month period. We extract those top users’ historical tweets using Twitter User Timeline API, without applying a spatial boundary limit. This method has a maximum number of tweets that can be collected from a specified user, producing varied time spans and varied tweet numbers, as not all users reached the 3200-tweet maximum.

Preprocessing and statistics of Twitter data

All the geotagged tweets are preprocessed to reduce potential artefacts causing biases in travel demand estimation. First, we only keep tweets that were generated from mobile devices. Moreover, those users who only had geotagged tweets of a single place are removed due to being bot accounts, e.g., for job posting or weather updates (Ek and Wennerberg 2020 ). Next, Twitter users can cross-post geotagged tweets from other social media platforms, yielding a place’s location being posted instead of the tweet’s precise geolocation, for example, the centre of Sweden or the centre of Gothenburg. These geotagged tweets without precise GPS coordinates are also removed. Finally, two filters are implemented for Twitter LD only. The top geotag Twitter users who have less than 50 geotagged tweets in total are removed. Considering the long time span of a given Twitter user’s Twitter timeline, he/she might have migrated from one country to another. To avoid confusion, we only keep the latest time period of the geotagged tweets where a Twitter user is assumed to live in Sweden. For the national level, all the Twitter LT and LD are used while for the city level, only these geotagged tweets within the boundary of Greater Gothenburg are used.

We derive the home and workplace locations from Twitter LD given the larger numbers of geotagged tweets per user. The home location is identified as the most-visited location on weekends and between 7pm and 8am on weekdays, whereas the most visited non-home location between 8am and 8pm on weekdays is identified as the user’s workplace (Schneider et al. 2013 ; Wang et al. 2018 ; Osorio-Arjona and García-Palomares 2019 ).

Following the practice in the literature to account for the fact that Twitter users are not representative of the overall population, we give weights for individual Twitter users in Twitter LD. The weight is the ratio of Twitter users to the true population in the municipality (Wang et al. 2018 ). The trips of the Twitter users in Twitter LD are aggregated and multiplied with their individual weight to derive a population-level travel demand estimation. The Twitter users’ distributions are found to correlate with the census (Kendall’s tau = 0.65, \(p<0.001\) ). However, top Twitter users tend to be over-represented in big cities especially the top three cities in Sweden: Stockholm (Twitter = 18% vs. Census = 9.4%), Gothenburg (6.9% vs. 5.6%), and Malmö (4.5% vs. 3.3%).

The basic statistics of Twitter LT and LD are summarised in Table 1 . Compared with Twitter LT, Twitter LD collected from user timelines without using any spatial bounding box covers a longer time span, contains a larger volume of geotagged tweets and a higher number of geotagged tweets per user and in total, but covers a smaller population than Twitter LT. The distribution of the number of total geotagged tweets per user is shown in Fig. 2 .

figure 2

Distribution of the total number of geotagged tweets of Twitter users by spatial scale and dataset

Swedish national travel survey (Survey)

The survey data come from the Swedish National Travel Survey (one-day travel diary) for the years of 2011 to 2016 (Official Statistics of Sweden 2016 ). It consists of a total of 171,553 trips from 38,258 participants covering 2189 record days, with detailed information on individual trip’s origin and destination, distance, travel time, and participant’s home/workplace. The spatial accuracy is the municipality level.

Model-based travel demand estimations (Sampers)

The Swedish Transport Administration uses the Sampers model to calculate changes in traffic volumes under different scenarios. Both the city level and the national level have their own traffic analysis zones that follow the census boundaries and homogeneous socioeconomic characteristics. These spatial zones are used for creating OD matrices with Twitter data so that we can compare Twitter with Sampers’ model output.

Sampers calculates travel demand based on studies of travel habits derived from travel surveys, looking at where, how and how often people want to travel, which forms the OD matrices. The model output represents the total travel demand for an average weekday. We used the latest OD matrices (2014) from Sampers for Greater Gothenburg and the entire Sweden. At the national level, we focus on the long-distance trips of Sampers model ( \(\ge\) 100 km).

Methodology

In order to examine the feasibility of using Twitter data for travel demand estimation, we use an analytic framework to compare Twitter with the other established data sources, as shown in Fig. 3 . In practice, transport planners collect empirical trip data from a small sample of the population and create a model to simulate the travel demand of the overall population for further application, such as traffic flows modelling. Therefore, we divide the comparison into two focuses: empirical trip records (“ Trip records ” section) and model output (“ Travel demand model construction ” section).

We first compare the empirical trip records obtained from Twitter with those from travel survey data with respect to the overall travel demand for an average weekday (“ Processing weekday trips ” section) and commuting travel demand (“ Processing commuting trips ” section). In this part of the validation, we also examine the stability of the similarity between Twitter and the travel survey over time. After the analysis of the empirical trips, we create the gravity models, based on Twitter data collected with two sampling methods, to simulate the overall travel demand at both the national (long-distance travel above 100 km) and city level. We use two methods for the step of trip generation (“ Trip generation ” section) followed by the gravity model for the trip distribution (“ Trip distribution ” section); they are trips converted from displacements by adding a time threshold ( Model A ) and the density-based approach proposed in this study ( Model B ). Model B is proposed as an alternative to Model A to solve the sparsity issue of Twitter data. Finally, we evaluate the results (“ Evaluation of Twitter OD matrices ” section) by comparing the Twitter-based trips and model outcomes with those from the national travel survey (Survey) and the Sampers model. The techniques used for the comparison include visualisation, similarity measure (“ Spatially weighted structural similarity index ” section), and trip distance distribution.

figure 3

The methodology used for examining the feasibility of Twitter data for travel demand modelling. Twitter LD is described in “ Longitudinal geotagged tweets (Twitter LD) ” section. Twitter LT is described in “ Lateral geotagged tweets (Twitter LT) ” section. Survey is described in “ Swedish National Travel Survey (Survey) ” section. Sampers is described in “ Model-based travel demand estimations (Sampers) ” section

Trip records

Processing weekday trips.

We define geotagged displacements by connecting every two consecutive geotagged tweets generated by the same user. To convert these displacements into trips, a time threshold can be used to filter out those displacements that have a time interval longer than a predefined threshold (Gao et al. 2014 ; Kheiri et al. 2015 ; Lee et al. 2019 ). We select 270 min, i.e., the 99th percentile of travel time between municipalities from Survey, as the time threshold for the national-level trip generation. For the city-level trip generation, the time threshold of 140 min is selected which is the 99th percentile of travel time within the corresponding county where most parts of Greater Gothenburg are located.

Survey contains complete sets of trip records at the municipality level. By directly aggregating the weekday records, we get the OD matrix of the overall trip records for an average weekday. The low cost of collecting Twitter data makes it easier to keep them updated over time. However, their actual use also depends on the stability of the similarity between Twitter trips and Survey trips over time. Instead of aggregating the records available, we look into the similarity of OD matrices from 2011 to 2016 at the national level by aggregating the records yearly.

Processing commuting trips

To construct commuting flows with Twitter LD, we define trips connecting home (origin) and workplace (destination), and aggregate those trips at the municipality level. This gives the national commuting OD matrix based on Twitter LD.

To compare a Twitter-based commuting OD matrix we need to construct an equivalent Survey-based OD matrix. Survey has the home and workplace of each participant at the municipality level, and each participant is assigned an individual weight standing for the representativeness of his/her socio-demographic profile in the overall Swedish population, regarding the time period of participating the survey, region, age, and gender. Specifically, the weight is designed as the ratio between the population and the survey respondent in the respective stratum. By linking home and workplace as a commuting trip for a given individual, multiplied by his/her individual weight, we aggregate all the commuting trips and construct the national commuting OD matrix.

Travel demand model construction

This section introduces the method of taking empirical trips to create modelled output of travel demand (OD matrix). The method consists of two steps, trip generation (“ Trip generation ” section) and trip distribution (“ Trip distribution ” section).

Displacement conversion (Model A)

By aggregating all the trips converted from displacements (see “ Processing weekday trips ” section), we get the overall OD matrices for an average weekday at the national level and the city level.

Based on the OD matrix that is aggregated directly from the trip records, the productions, \(P_i\) for a given origin zone i , are expressed as the summation of \(f_{ij}\) , i.e., the number of trips between the origin zone i and the destination zone j , over the destination zone j ( \(P_i=\sum _{j=1}^{N}f_{ij},i=1,2,\ldots ,N\) ), where N is the total number of zones. Similarly, trip attractions are expressed as \(A_j^0=\sum _{i=1}^{N}f_{ij},j=1,2,\ldots ,N\) .

Density-based approach (Model B)

Given the sparsity of Twitter data, the reduction of available data due to adding a time threshold further limits the use of geotagged tweets to represent the travel demand at higher granularity. Therefore, we propose an alternative way utilising all the available geotagged tweets and the census data.

The population of each zone ( \(\text {Pop}\) ) represents its productions, i.e., \(P_i=\text {Pop}_i,i=1,2,\ldots ,N\) . And the attractions are represented by the number of geotagged tweets in each zone ( \(f_j\) ), i.e., \(A_j^0=f_{j},j=1,2,\ldots ,N\) .

For both methods, the trip productions and attractions are balanced so that they sum up to the same number.

The gravity model was first proposed for the estimation of an OD matrix in the 1940s (Zipf 1946 ) and later became one of the most applied methods for the trip distribution (Yang et al. 2015 ). Given that this study focuses on the data source, we select the following form of gravity model to avoid model complexity:

where \(T_{ij}\) is the number of trips between the origin zone i and the destination zone j , \(F_{ij}\) is the friction factor for travelling between zone i and j . \(F_{ij}\) is defined below:

where \(d_{ij}\) is the Haversine distance between the centroid of zone i and zone j . \(\alpha\) is set to 1. And \(\beta\) is calibrated with the OD matrix directly derived from the raw data so that they are optimally approximated by the estimated OD matrix where the similarity is measured by SpSSIM. Trip distribution uses Iterative Proportional Fitting (IPF) to assign trips from the predefined productions and attractions to the estimated OD matrix (Ben-Akiva et al. 1985 ; McCord et al. 2010 ). All the OD matrices are standardised, so that every cell has a value between 0 and 1 representing the probability of the connection between two zones.

Evaluation of Twitter OD matrices

The comparison techniques used to evaluate the Twitter OD matrices include the visualisation of the OD matrices and the similarity measure (SpSSIM) between the OD matrix from Twitter and from the external sources (“ Spatially weighted structural similarity index ” section). An essential aspect of human mobility behaviour is the travel distance ( d , km) of OD pairs, whose distribution provides another facet for validating Twitter to estimate travel demand. Therefore, we compare this distribution from Twitter data with that from other sources whose similarity is measured by Kullback-Leibler (KL) divergence measure (Wang et al. 2019 ; Smolak et al. 2020 ). The smaller the KL divergence, the more similar are the two given distributions.

Spatially weighted structural similarity index

To evaluate the feasibility of using Twitter data as a proxy for travel demand, we compare Twitter-based OD matrices with those from external data sources and measure their similarity. The more similar, the better Twitter data work as a source for travel demand estimation.

Originally proposed by Wang et al. ( 2004 ), the structural similarity (SSIM) measures the similarity between two images for assessing image quality. We use the spatially weighted structural similarity index (SpSSIM) (Jin et al. 2019 ) that was later introduced to the field of transport for comparing the quality of OD matrices that are based on different data sources (Djukic et al. 2013 ; Pollard et al. 2013 ). This newly proposed SpSSIM overcomes the SSIM’s sensitivity issue due to the ordering of OD pairs, as raised by earlier studies (e.g., Djukic 2014 ).

The distances between all possible OD pairs create a matrix \(\mathbf {D}\) where \(d_{ij}\) is the Haversine distance between the centroids of zone i and zone j . The distances are binned by their percentile, yielding multiple groups of the same number of OD pairs (10% per group) based on their spatial adjacency ( \([D_{min}^b,D_{max}^b],b=1,2,\ldots ,n\) ). The spatial filtering matrix ( \(\mathbf {W}^b\) ) consists of 0 and 1, and is defined as:

Two OD matrices, \(\mathbf {X}\) and \(\mathbf {Y}\) , are in a probabilistic form i.e., each cell is a value between 0 and 1 indicating the strength/probability of the connection between zone i and zone j . For a distance group, all the cells beyond the range of the distance are set to zero, with \(\mathbf {W}^{b}\mathbf {X}\) and \(\mathbf {W}^{b}\mathbf {Y}\) indicating the Hadamard product between \(\mathbf {W}^{b}\) and \(\mathbf {X}\) and \(\mathbf {W}^{b}\) and \(\mathbf {Y}\) , respectively. The similarity between \(\mathbf {X}\) and \(\mathbf {Y}\) on this distance range is indicated by \(\text {SpSSIM}\left( \mathbf {X}, \mathbf {Y},\mathbf {W}^{b}\right)\) which is:

where \(\mu\) is the mean, \(\sigma\) the variance or the covariance between two matrices, and \(C_1\) and \(C_2\) are two constants.

The share of the travel demand for a distance group b is expressed below:

where N indicates the number of traffic zones and \(s^b\) has a value between 0 and 1 indicating the share of trips that are expected to happen within the distance range b . The OD pairs of different distance groups have imbalanced share of travel demand between them ( \(s^b,b=1,2,\ldots ,n\) ). Accounting for the share of travel demand, the similarity between the two OD matrices ( \(\mathbf {X}\) and \(\mathbf {Y}\) ) is quantified by SpSSIM as calculated below aggregating over all the distance groups:

For the matrices in this study, have a mean squared \(\mu ^2\sim 10^{-12}\) – \(10^{-9}\) and \(\sigma \sim 10^{-9}\) – \(10^{-6}\) , hence we adjust the constants to values of \(C_1=10^{-16}\) – \(10^{-13}\) , \(C_2=10^{-11}-10^{-8}\) . The detailed justification of these selections can be found in a study by Pollard et al. ( 2013 ). From the above definition, SpSSIM has a value between 0 and 1. SpSSIM equals to 1 when two OD matrices have the same exact pattern.

Trip records comparison

Commuting trips.

The commuting OD matrices of Survey and Twitter LD are shown in Fig. 4 . The diagonal cells have higher values indicating that most individuals commute within their residence municipality. Spatial proximity between municipalities (neighbouring cells) also affects the inter-municipality commuting flows in Survey’s OD matrix. However, this is not the same for the OD matrix derived from Twitter LD: the estimated home and workplace are not strongly influenced by the distance between them. The similarity between Twitter LD and Survey is low (SpSSIM = 0.39, KL divergence = 0.052).

figure 4

Commuting origin (y-axis) and destination (x-axis) OD matrices based on Survey (left figure) and Twitter LD (right figure) trip records. Colour showing the probability of connections between each pair of zones, i.e., the proportion of trips. (Color figure online)

Commuting travel distances are shown in Fig. 5 a. The distribution produced by Twitter LD is significantly higher than Survey. This is consistent with the observation in Fig. 4 that Twitter LD does not capture the frequent commuting flows between neighbouring cells as shown in Survey’s OD matrix.

figure 5

Travel distance distribution of data for Twitter and Survey. Cumulative share of trips is the probability of travelling between zones at a distance equal or below a certain threshold. a Commuting trips. b Weekday trips. (Color figure online)

Weekday trips

The weekday trips’ OD matrices (Fig. 6 ) from the two Twitter datasets and Survey are visually more similar than the results of commuting travel demand shown in “ Commuting trips ” section. Twitter data produce sparse matrices, and Twitter LT is sparser than Twitter LD. In addition, Twitter LD looks more similar to Survey as compared to Twitter LT.

The quantitative similarity (SpSSIM) results are in line with the visual results; at the national level, the trips converted from displacements in Twitter LD approximate the Survey OD matrix better than those from Twitter LT (0.87 vs. 0.64). Still at the national level, the Twitter trips collected with either sampling method approximate the Survey’s trip distance distribution well (Fig. 5 b) in contrast with a greater discrepancy observed in commuting travel distances (Fig. 5 a).

figure 6

OD matrices based on the weekday trip records from Survey and Twitter data

The similarity of trips disaggregated by year is rather stable, when comparing Survey to the baseline year (2011), and Twitter LD to the Survey over time (see Fig. 7 ). The stability of this similarity between Twitter LD and Survey suggests that Twitter data are reliable in capturing changes in travel and thus are suitable for estimating the change in national-level travel demand over time.

figure 7

Similarity between Survey and Twitter LD and its sample size by year. a Survey vs. Twitter LD over time. The curve of Survey shows how the OD matrix deviates from the baseline year, 2011 for Survey records. b Number of geotagged tweets in Twitter LD over time

Model outcomes

When looking at the similarity with Sampers, Twitter data generally work well at the city level (0.54 to 0.85), while the performance at the national level is not as good (0.40 to 0.54), see Table 2 . The sampling method matters; Twitter LD is more similar to Sampers than Twitter LT, especially when using Model A with Displacement conversion (Twitter \(\hbox {LD}^A\) vs. Twitter \(\hbox {LT}^A\) , National = 0.52 vs. 0.40, and City = 0.74 vs. 0.54). Combining the density-based approach and the gravity model (Model B) produces better similarity results compared to using Displacement conversion (Model A) at both spatial scales. This is probably due to the fact that Model B manages to increase the number of available geotagged tweets five-fold relative to Model A.

Figure 8 shows model outcomes. At both the national level (upper row) and city level (bottom row), the visualisation confirms the similarity as quantified by the SpSSIM and KL divergence values shown in Table 2 . Comparing the two spatial scales, the greater number of traffic zones and larger geographical coverage make the national level more challenging to model using Twitter data due to the sparsity issue, leading to lower values of similarity in general than the city level.

figure 8

Estimated OD matrices by gravity model and Sampers’ model outputs. A: Displacement conversion plus gravity model. B: Density-based approach plus gravity model

Finally, we look into the travel distance distribution (Fig. 9 ). At the national level, the Twitter-based output with Model A shows greater short-distance travel demand when compared with the Sampers model (Fig. 9 a, b). At the city level, however, the Twitter models approximate the traffic model better than at the national level in general, especially for Twitter LD (Fig. 9 c, d). At both spatial scales, Model B represents the trip distance distribution better than Model A.

figure 9

Trip distance distribution. Cumulative share of trips refers to the probability of travel between zones below a given distance. The trip distance is from the estimated OD matrices by A displacement conversion plus gravity model and by B density-based approach plus gravity model. a National level—Twitter LD. b National level—Twitter LT. c City level—Twitter LD. d City level—Twitter LT

Sensitivity of Twitter-based model outcomes to the sample size

Sensitivity of outcomes to sample size and to sampling method of tweets (LD or LT) are tested using a share of geotagged tweets from 1% to 99%, with a step length of 1% and 10 repetitions of random sampling, to create outputs using models A and B with the same settings as above. Figure 10 shows the similarity results.

figure 10

Similarity, a SpSSIM and b KL divergence, as a function of the number of geotagged tweets. Green colours show the results using Twitter LT and blue colours show the results using Twitter LD. For the 10 model runs of each tweets sample size, the curve shows the average value of SpSSIM/KL divergence and the shaded area shows the maximum and minimum value of SpSSIM/KL divergence. Model A—displacement conversion plus gravity model; Model B—density-based approach plus gravity model. For all models, \(\beta = 0.03\) . (Color figure online)

As expected, as more geotagged tweets are included in the modelling, the similarity between the outputs of the Twitter-based OD matrix and the Sampers model increases and remains within a smaller range. The national level is more sensitive to data sparsity, because the number and the geographical coverage of traffic zones is greater than at the city level, therefore requires a greater number of tweets to reach a stable (but still lower) similarity. In terms of methodology, Model A is more sensitive to the number of geotagged tweets than Model B, especially with respect to the stability of the results with a smaller number of tweets, and is generally associated with poorer results.

Based on the comparison with travel surveys and the government’s traffic simulation model, our study suggests that geotagged tweets can be suitable for estimating the overall travel demand (including OD matrix and travel distance) for an average weekday. However, as discussed in “ Commuting travel demand estimation ” section, estimation of the commuting travel demand is not reliable even though the data have been used for this purpose in the literature (e.g., Zagatti et al. 2018 ). We further discuss the impact of spatial scale (“ The impact of spatial scale ” section), sampling size and sampling method (“ The impact of sampling size and methods of data collection ” section). In “ A novel density-based approach: geotagged tweets as attractions generators as opposed to trips generators ” section, we discuss the clear advantage of the innovative density-based approach proposed here and offer possible explanations for this result.

The reliability of estimated commuting trips using geotagged tweets is low.

We use a simple, yet commonly adopted method (Wang et al. 2018 ) to identify workplace locations: the most visited non-home location during 7 am–8 pm on weekdays. However, not all Twitter users are employed (according to the OECD, employment rate is 77.1% in Sweden for those aged 15–64) and not all work outside of home between 8 am–8 pm. Despite the difference between Twitter users and the general population, one can expect that the aggregate commuting trips should be quite similar given our findings with regards to overall trips. The observed dissimilarity could be explained by the errors introduced using the simple method mentioned above, and the behaviour biases of Twitter users. Most Twitter users may not feel comfortable or interested in geolocating their homes and workplaces online publicly due to privacy concerns. A geotag usage survey based on 400 US residents shows that 70% of their geotags happen in places that people visit infrequently (Tasse et al. 2017 ). The temporal distribution of geotagging behaviour resembles that of a leisure activity pattern (Federal Office for Spatial Development ARE 2017 ). Moreover, geotag users tend to geotag locations that are not within their neighbourhood; and the geotagged locations concentrate substantially at locations farther away from the daily mobility area (Tasse et al. 2017 ). This suggests Twitter can have significant shortcomings when used for capturing routine activities such as trips between home and the workplace.

The impact of spatial scale

The main obstacle of using Twitter data at a large spatial scale is the sparsity.

Using geotagged tweets for travel demand estimation requires a sufficient sample size, which depends on the form of Twitter data (“ The impact of sampling size and methods of data collection ” section), penetration rate, and the number of samples collected. An important contribution of our work is to examine the robustness of the travel demand estimation for different spatial scales. Most studies have focused on one particular scale, be it at the city (Wang et al. 2018 ) or international (Hawelka et al. 2014 ) level. Yet, the validity of the selection of scale for the purpose of travel demand estimation remains unclear until different spatial scales are properly investigated. Despite the accumulation of geotagged tweets over months (Twitter LT) to years (Twitter LD), the share of zones with insufficient coverage increases at the national level, due to Twitter’s lower penetration rate outside urban centres. Therefore, using geotagged tweets for travel demand estimation requires appropriate selection of spatial aggregation i.e., zoning system.

The impact of sampling size and methods of data collection

The more geotagged tweets included in the modelling, the better Twitter is at estimating travel demand. Twitter LD results in a much larger number of geotagged tweets that overall better represents population mobility patterns.

The number of geotagged tweets in Twitter LT is 30% of those in Twitter LD (3.4 vs. 11.5 million geotagged tweets). Compared with Twitter LT, Twitter LD covers longer period (9 years compared with 6 months for Twitter LT) with fewer users (2311 compared with 24,442 with Twitter LT). Our study demonstrates that, however, the long-term coverage of longitudinal geotagged tweets by top users (User Timeline API) compensates for the time sparsity and helps to recreate a more complete picture of population mobility patterns, and therefore, is more reliable for travel demand estimation than the lateral dataset (Twitter LT). However, this gap narrows or disappears when using a novel density-based approach developed in this study, see below in “ A novel density-based approach: geotagged tweets as attractions generators as opposed to trips generators ” section.

A novel density-based approach: geotagged tweets as attractions generators as opposed to trips generators

The density-based approach utilises more geotagged tweets, resulting in better representation of travel demand.

The common practice of adding a time threshold (Model A, displacement method) to capture “trips” drastically reduces the available Twitter data for travel demand estimation: only 20–35% of geotagged tweets are utilised to estimate the overall travel demand. This reduction limits the application of geotagged tweets given that sparsity is already one of its key drawbacks.

Without the need for a time threshold, the density-based approach (Model B) increases usable data by 2–7 times. This drastically increases the similarity scores of the OD matrices compared with Sampers’ model outputs and the method is not so sensitive to sample size. Considering both similarity and stability in Fig. 10 , a magnitude of 1000 geotagged tweets is sufficient for the city-level and the national level requires 10,000 tweets to reach a stable similarity. This is equivalent to a minimum of 1 geotagged tweet per 1000 persons for the entire Sweden and 2 geotagged tweets per 1000 persons for the more densely populated city, requiring roughly an order of magnitude fewer samples than the other method.

Not only does the density-based approach produce better OD matrices, it also produces better trip distance distributions compared with the Surveys. Tasse et al. ( 2017 ) suggest that most Twitter users geotag their tweets within an hour of arrival (if at all), thus geotagging may be a timely indicator of the start time of the activity. This emphasises that Twitter users geotweet to report activities instead of trips, therefore, the density of geotagged tweets naturally reflects the attractiveness of zones. This motivates the proposed density-based approach, which regards the tweets density of zones as the attractions and the population size of zones as the productions. The results suggest that the density-based approach captures some of the population flows, because we assume that the generated trips between zones are proportional to (1) the population and (2) the number of activities some of which are geotagged.

A plausible explanation could be that the improvement was solely ascribed to the use of population as production, instead of the use of geotagged tweets as attractions. To test this assumption, we observe the changes in similarity metric when we assign attraction and production using: (1) population count as both production and attraction and (2) geotagged tweets count as both production and attraction. They both perform better than the displacement conversion; however, they are not as good as the density-based approach.

The density-based approach can be extended to compute time-dependent attractions by aggregating geotagged tweets across different temporal profiles, providing a dynamic picture of travel demand by time of day, week, or season.

This study critically examines the feasibility of using geolocations of Twitter data to estimate population mobility. The overall results suggest that Twitter data can be suitable for modelling the overall travel demand for an average weekday but not the commuting travel demand due to the low reliability of identifying home and work locations. This makes it hard to replace the conventional national travel surveys and similar survey methods, in which users report the purposes of all trips.

The key strengths of social media data are that they are low-cost, abundant, available in real-time, and free of arbitrary geographical partition. However, there are also significant limitations: population and behavioural biases and lack of important information such as social demographic information and trip purposes. Despite clear indications of overly representing residents in big cities and their leisure activities from the existing literature, we demonstrate in the present study that geotagged tweets can provide a reasonably good travel demand estimation that also captures the trends over time.

Limitations and outlook

The present study uses data from a single country. Further exploration is needed to understand how the findings can be generalised to the other regions, despite comparable validation data being difficult to come by. The proposed density-based approach allows flexible temporal aggregation to estimate time-varying travel demand, however, our validation data only represent an average day without any time-dependent demand estimation and exclude weekends and long-distance travel. This limits our ability to test the validity of time-dependent travel demand estimates and travel outside of the validation regions. One future direction is to test the performance of the density-based approach at different levels of temporal and spatial resolution. In addition, future work can use the tweet contents, not used in the current study, as this can provide additional information for inferring trip purposes. Last but not least, despite adjusting the Twitter users in Twitter LD based on its ratio to the true population at the municipality level, the method is simple and can be further improved, such as considering additional socio-demographic dimensions, to better represent the population. In the future, we can explore methods such as inferring trip purposes and socio-demographic information to derive more robust and more reliable travel demand estimation.

Abbasi, A., Rashidi, T.H., Maghrebi, M., Waller, S.T.: Utilising location based social media in travel survey methods: bringing twitter data into the play. In: Proceedings of the 8th ACM SIGSPATIAL international workshop on location-based social networks, pp. 1–9 (2015)

Anda, C., Erath, A., Fourie, P.J.: Transport modelling in the age of big data. Int. J. Urban Sci. 21 (sup1), 19–42 (2017)

Article   Google Scholar  

Barbosa, H., Barthelemy, M., Ghoshal, G., James, C.R., Lenormand, M., Louail, T., Menezes, R., Ramasco, J.J., Simini, F., Tomasini, M.: Human mobility: models and applications. Phys. Rep. 734 , 1–74 (2018)

Ben-Akiva, M., Macke, P.P., Hsu, P.S.: Alternative methods to estimate route-level trip tables and expand on-board surveys. Transp. Res. Record 1037 , 1 (1985)

Google Scholar  

Calabrese, F., DiLorenzo, G., Liu, L., Ratti, C.: Estimating origin–destination flows using opportunistically collected mobile phone location data from one million users in Boston metropolitan area. IEEE Pervasive Comput. 10 (4), 36–44 (2011)

Cheng, Z., Jian, S., Rashidi, T.H., Maghrebi, M., Waller, S.T.: Integrating household travel survey and social media data to improve the quality of od matrix: a comparative case study. IEEE Trans. Intell. Transp. Syst. 21 (6), 2628–2636 (2020)

Djukic, T.: Dynamic OD demand estimation and prediction for dynamic traffic management. TRAIL Thesis Series no. T2014/9, Delft University of Technology, 2600 GA Delft, The Netherlands (2014)

Djukic, T., Hoogendoorn, S., Van Lint, H.: Reliability assessment of dynamic od estimation methods based on structural similarity index. Technical Report (2013)

Ek, K., Wennerberg, E.: Estimating travel demand from twitter using an individual mobility model: in Sweden, the Netherlands and São paulo. Master’s thesis, Chalmers tekniska högskola, S-412 96 Gothenburg, Sweden, https://hdl.handle.net/20.500.12380/301742 (2020)

Federal Office for Spatial Development ARE: Population’s transport behaviour 2015. Technical Report (2017)

Gao, S., Yang, J.A., Yan, B., Hu, Y., Janowicz, K., McKenzie, G.: Detecting origin–destination mobility flows from geotagged tweets in greater Los Angeles area. In: Eighth International Conference on Geographic Information Science (GIScience’14), Citeseer (2014)

Gonzalez, M.C., Hidalgo, C.A., Barabasi, A.L.: Understanding individual human mobility patterns. Nature 453 (7196), 779–782 (2008)

Hasan, S., Ukkusuri, S.V.: Urban activity pattern classification using topic models from online geo-location data. Transp. Res. Part C: Emerg. Technol. 44 , 363–381 (2014)

Hasnat, M.M., Hasan, S.: Identifying tourists and analyzing spatial patterns of their destinations from location-based social media data. Transp. Res. Part C: Emerg. Technol. 96 , 38–54 (2018)

Hasnat, M.M., Faghih-Imani, A., Eluru, N., Hasan, S., et al.: Destination choice modeling using location-based social media data. J. Choice Modell. 31 , 22–34 (2019)

Hawelka, B., Sitko, I., Beinat, E., Sobolevsky, S., Kazakopoulos, P., Ratti, C.: Geo-located twitter as proxy for global mobility patterns. Cartogr. Geogr. Inf. Sci. 41 (3), 260–271 (2014)

Jeuken, G.S.: Using big data for human mobility patterns—examining how twitter data can be used in the study of human movement across space. Master’s thesis (2017)

Jin, P., Cebelak, M., Yang, F., Zhang, J., Walton, C., Ran, B.: Location-based social networking data: exploration into use of doubly constrained gravity model for origin–destination estimation. Transp. Res. Record J. Transp. Res. Board 2430 , 72–82 (2014)

Jin, C., Nara, A., Yang, J.A., Tsou, M.H.: Similarity measurement on human mobility data with spatially weighted structural similarity index (SpSSIM). Trans. GIS 24 (1), 104–22 (2020)

Jurdak, R., Zhao, K., Liu, J., AbouJaoude, M., Cameron, M., Newth, D.: Understanding human mobility from twitter. PloS One 10 (7), e0131469 (2015)

Kheiri, A., Karimipour, F., Forghani, M.: Intra-urban movement flow estimation using location based social networking data. Int. Arch. Photogr. Remote Sens. Spatial Inf. Sci. 40 (1), 785 (2015)

Lee, J.H., Gao, S., Goulias, K.G.: Can twitter data be used to validate travel demand models. In: 14th International Conference on Travel Behaviour Research (2015)

Lee, J.H., Davis, A., McBride, E., Goulias, K.G.: Statewide comparison of origin–destination matrices between California travel model and twitter. Mobility Patterns, Big Data and Transport Analytics, pp. 201–228. Elsevier, Amsterdam (2019)

Lenormand, M., Picornell, M., Cantú-Ros, O.G., Tugores, A., Louail, T., Herranz, R., Barthelemy, M., Frias-Martinez, E., Ramasco, J.J.: Cross-checking different sources of mobility information. PLoS One 9 (8), e105184 (2014)

Lenormand, M., Gonçalves, B., Tugores, A., Ramasco, J.J.: Human diffusion and city influence. J. R. Soc. Interface 12 (109), 20150473 (2015)

Liao, Y., Yeh, S.: Predictability in human mobility based on geographical-boundary-free and long-time social media data. In: 2018 21st International Conference on Intelligent Transportation Systems (ITSC). IEEE, pp. 2068–2073 (2018)

Liao, Y., Yeh, S., Jeuken, G.S.: From individual to collective behaviours: exploring population heterogeneity of human mobility based on social media data. EPJ Data Sci. 8 (1), 34 (2019)

Maghrebi, M., Abbasi, A., Rashidi, T.H., Waller, S.T.: Complementing travel diary surveys with twitter data: application of text mining techniques on activity location, type and time. In: 2015 IEEE 18th international conference on intelligent transportation systems (ITSC). IEEE, pp 208–213 (2015)

McCord, M.R., Mishalani, R.G., Goel, P., Strohl, B.: Iterative proportional fitting procedure to determine bus route passenger origin–destination flows. Transp. Res. Record 2145 (1), 59–65 (2010)

McNally, M.G.: The four-step model. In: Hensher, D.A., Button, K.J. (eds.) Handbook of Transport Modelling, vol. 1, pp. 35–53. Emerald Group Publishing Limited (2007). https://doi.org/10.1108/9780857245670-003

Mislove, A., Lehmann, S., Ahn, Y.Y., Onnela, J.P., Rosenquist, J.N.: Understanding the demographics of twitter users. ICWSM 11 (5), 25 (2011)

Molloy, J., Moeckel, R.: Improving destination choice modeling using location-based big data. ISPRS Int. J. Geo-Inf. 6 (9), 291 (2017)

Morstatter, F., Pfeffer, J., Liu, H., Carley, K.M.: Is the sample good enough? Comparing data from twitter’s streaming API with twitter’s firehose. In: ICWSM (2013)

Official Statistics of Sweden (2016) Swedish National Travel survey (RVU Sweden) 2011–2016. https://www.trafa.se/en/travel-survey/travel-survey/

Osorio-Arjona, J., García-Palomares, J.C.: Social media and urban mobility: using twitter to calculate home-work travel matrices. Cities 89 , 268–280 (2019)

Pollard, T., Taylor, N., van Vuren, T., MacDonald, M.: Comparing the quality of OD matrices in time and between data sources. In: Proceedings of the European Transport Conference (2013)

Rashidi, T.H., Abbasi, A., Maghrebi, M., Hasan, S., Waller, T.S.: Exploring the capacity of social media data for modelling travel behaviour: opportunities and challenges. Transp. Res. Part C: Emerg. Technol. 75 , 197–211 (2017)

Ribeiro, A.I.J.T., Silva, T.H., Duarte-Figueiredo, F., Loureiro, A.A.: Studying traffic conditions by analyzing foursquare and instagram data. In: Proceedings of the 11th ACM symposium on Performance evaluation of wireless ad hoc, sensor, and ubiquitous network. ACM, pp. 17–24 (2014)

Schneider, C.M., Belik, V., Couronné, T., Smoreda, Z., González, M.C.: Unravelling daily human mobility motifs. J. R. Soc. Interface 10 (84), 20130246 (2013)

Smolak, K., Rohm, W., Knop, K., Siła-Nowicka, K.: Population mobility modelling for mobility data simulation. Comput. Environ. Urban Syst. 84 , 101526 (2020)

Song, C., Qu, Z., Blumm, N., Barabási, A.L.: Limits of predictability in human mobility. Science 327 (5968), 1018–1021 (2010)

Tasse, D., Liu, Z., Sciuto, A., Hong, J.I.: State of the geotags: Motivations and recent changes. In: ICWSM, pp. 250–259 (2017)

Twitter, Inc.: Filter realtime tweets. https://developer.twitter.com/en/docs/tweets/filter-realtime/api-reference/post-statuses-filter (2019a)

Twitter, Inc.: Get Tweet timelines. https://developer.twitter.com/en/docs/tweets/timelines/overview (2019b)

Twitter, Inc.: Twitter provides Tweets and associated metadata including geo data, images, and mentions. https://support.gnip.com/sources/twitter/ (2019c)

Wang, Z., Bovik, A.C., Sheikh, H.R., Simoncelli, E.P., et al.: Image quality assessment: from error visibility to structural similarity. IEEE Trans. Image Process. 13 (4), 600–612 (2004)

Wang, Q., Phillips, N.E., Small, M.L., Sampson, R.J.: Urban mobility and neighborhood isolation in America’s 50 largest cities. Proc. Natl. Acad. Sci. 115 (30), 7735–7740 (2018)

Wang, J.X., Huang, J., Duan, L., Xiao, H.: Prediction of Reynolds stresses in high-mach-number turbulent boundary layers using physics-informed machine learning. Theor. Comput. Fluid Dyn. 33 (1), 1–19 (2019)

Wesolowski, A., Eagle, N., Noor, A.M., Snow, R.W., Buckee, C.O.: The impact of biases in mobile phone ownership on estimates of human mobility. J. R. Soc. Interface 10 (81), 20120986 (2013)

Yang, F., Jin, P.J., Cheng, Y., Zhang, J., Ran, B.: Origin–destination estimation for non-commuting trips using location-based social networking data. Int. J. Sustain. Transp. 9 (8), 551–564 (2015)

Yue, Y., Lan, T., Yeh, A.G., Li, Q.Q.: Zooming into individuals to understand the collective: a review of trajectory-based travel behaviour studies. Travel Behav. Soc. 1 (2), 69–78 (2014)

Zagatti, G.A., Gonzalez, M., Avner, P., Lozano-Gracia, N., Brooks, C.J., Albert, M., Gray, J., Antos, S.E., Burci, P., zu Erbach-Schoenberg, E., et al.: A trip to work: estimation of origin and destination of commuting patterns in the main metropolitan regions of Haiti using CDR. Dev. Eng. 3 , 133–165 (2018)

Zhang, Z., He, Q., Zhu, S.: Potentials of using social media to infer the longitudinal travel behavior: a sequential model-based clustering method. Transp. Res. Part C: Emerg. Technol. 85 , 396–414 (2017)

Zipf, G.K.: The p 1 p 2/d hypothesis: on the intercity movement of persons. Am. Soc. Rev. 11 (6), 677–686 (1946)

Download references

Open Access funding provided by Chalmers University of Technology. The authors acknowledge the financial support of the Swedish Research Council for Sustainable Development (Formas, project number 2016-01326).

Author information

Authors and affiliations.

Division of Physical Resource Theory, Department of Space, Earth and Environment, Chalmers University of Technology, Gothenburg, Sweden

Yuan Liao & Sonia Yeh

Division of Urban Design and Planning, Department of Architecture and Civil Engineering, Chalmers University of Technology, Gothenburg, Sweden

You can also search for this author in PubMed   Google Scholar

Contributions

Yuan Liao, Sonia Yeh and Jorge Gil conceptualised the study. All the authors designed the methods. Yuan Liao analysed the data. All the authors wrote the manuscript.

Corresponding author

Correspondence to Yuan Liao .

Ethics declarations

Conflicts of interest.

The authors declare no competing financial or non-financial interests.

Availability of data and material

The aggregate data that support the findings of this study are available upon request from the corresponding author. The individual data can not be made available since they contain information that could compromise the privacy of individuals.

Code availability

Not applicable.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Liao, Y., Yeh, S. & Gil, J. Feasibility of estimating travel demand using geolocations of social media data. Transportation 49 , 137–161 (2022). https://doi.org/10.1007/s11116-021-10171-x

Download citation

Accepted : 11 January 2021

Published : 26 January 2021

Issue Date : February 2022

DOI : https://doi.org/10.1007/s11116-021-10171-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

  • Social media data
  • Travel demand
  • Origin–destination estimation
  • Longitudinal data
  • Lateral data
  • Gravity model
  • Find a journal
  • Publish with us
  • Track your research

US Travel Header Utility Menu

  • Future of Travel Mobility
  • Travel Action Network
  • Commission on Seamless & Secure Travel
  • Travel Works
  • Journey to Clean

Header Utility Social Links

  • Follow us on FOLLOW US
  • Follow us on Twitter
  • Follow us on LinkedIn
  • Follow us on Instagram
  • Follow us on Facebook

User account menu

The latest travel data.

MONTHLY INSIGHTS March 04, 2024

travel social data

U.S. Travel has temporarily paused our monthly data newsletter, however, the latest travel data is still available via the U.S. Travel Insights Dashboard . This dashboard is updated each month (member login required).

The U.S. Travel Insights Dashboard , developed in collaboration with Tourism Economics, is supported by more than 20 data sources. The dashboard is the most comprehensive and centralized source for high-frequency intelligence on the U.S. travel industry, tracking industry performance, travel volumes and predictive indicators of recovery including air and lodging forecasts, DMO website traffic, convention and group trends, travel spending and losses, traveler sentiment, among others to measure the health of the industry.

Key Highlights January 2024:

  • Travel appetite started the year on a softer note, but overall growth continued. Air passenger growth remained positive, up 6% versus the prior year but lower than the double-digit growth seen through 2023. Foreign visits remained strong, up 24% YoY.
  • Hotel room demand continued a trend of slight contraction falling 1% versus the prior year, while short-term rental demand grew 1%, a lower rate than 2023.
  • A particular bright spot was that group room demand within the top 25 markets displayed solid growth of 9% relative to the prior year.
  • The outlook for the economy remains fairly optimistic due to the strength of the labor market, looser financial conditions and healthy household and nonfinancial corporate balance sheets. This has filtered through to slightly higher consumer sentiment in February.
  • Sentiment is also growing for upcoming leisure travel in 2024. The share of travelers reporting having travel plans within the next six months increased to 93% in January from 92% in December, according to Longwoods International’s monthly survey.
  • Travel price inflation (TPI) fell slightly in January as a result of falling transportation prices. Sticky services inflation should see relief from decelerating wage growth. However, upside risks stem from rising healthcare costs, supply chain disruptions and slowing labor supply. Source: U.S. Travel Association and Tourism Economics

Member Price:

Non-Member Price: Become a member to access.

ADDITIONAL RESEARCH

Travel Price Index

Travel Forecast

Quarterly Consumer Insights

Additional monthly insights are available through the full U.S. Travel Monthly Data Report, exclusive to members. Please inquire with  membership  if you are interested in learning about becoming a member of U.S. Travel Association. 

FinanceBuzz

FinanceBuzz

12 Ways To Keep Your Personal Data Safe While Traveling

Posted: January 15, 2024 | Last updated: January 15, 2024

<p>Whether you're a seasoned globetrotter or a weekend warrior, every adventure comes with a hidden risk: the threat to your digital security. </p><p>While pickpockets and lost passports are travel woes of old, modern dangers lurk in the shadows of Wi-Fi hotspots and public charging stations. </p><p>So, <a href="https://financebuzz.com/ways-to-travel-more?utm_source=msn&utm_medium=feed&synd_slide=1&synd_postid=15545&synd_backlink_title=step+up+your+travel+game&synd_backlink_position=1&synd_slug=ways-to-travel-more">step up your travel game</a> with these essential cybersecurity tips and keep your data safe wherever your journeys take you.</p> <p>  <a href="https://financebuzz.com/top-travel-credit-cards?utm_source=msn&utm_medium=feed&synd_slide=1&synd_postid=15545&synd_backlink_title=Earn+Points+and+Miles%3A+Find+the+best+travel+credit+card+for+nearly+free+travel&synd_backlink_position=2&synd_slug=top-travel-credit-cards"><b>Earn Points and Miles:</b> Find the best travel credit card for nearly free travel</a>  </p>

Whether you're a seasoned globetrotter or a weekend warrior, every adventure comes with a hidden risk: the threat to your digital security.

While pickpockets and lost passports are travel woes of old, modern dangers lurk in the shadows of Wi-Fi hotspots and public charging stations.

So, step up your travel game with these essential cybersecurity tips and keep your data safe wherever your journeys take you.

Earn Points and Miles: Find the best travel credit card for nearly free travel

<p> If you’re staying in a hotel while traveling, your internet connection likely won’t be secure. That gives hackers an easy way to steal your data. </p> <p> One of the best ways to protect your data in these situations is by using a virtual private network or VPN. This is a great way to encrypt data and hide your IP address from prying eyes.  </p> <p>  <p class=""><a href="https://financebuzz.com/extra-newsletter-signup-testimonials-synd?utm_source=msn&utm_medium=feed&synd_slide=2&synd_postid=15545&synd_backlink_title=Get+expert+advice+on+making+more+money+-+sent+straight+to+your+inbox.&synd_backlink_position=3&synd_slug=extra-newsletter-signup-testimonials-synd">Get expert advice on making more money - sent straight to your inbox.</a></p>  </p>

If you’re staying in a hotel while traveling, your internet connection likely won’t be secure. That gives hackers an easy way to steal your data.

One of the best ways to protect your data in these situations is by using a virtual private network or VPN. This is a great way to encrypt data and hide your IP address from prying eyes.

Get expert advice on making more money - sent straight to your inbox.

<p> When you are at home, the odds are good that you select the auto-connect option for your Wi-Fi and internet connection for convenience. However, leaving this setting activated means you might accidentally connect to a nefarious network when traveling.  </p> <p> So, before you travel, turn off this feature on your phone, laptop, tablet, and other electronic devices, just to be safe.  </p>

Disable auto-connect

When you are at home, the odds are good that you select the auto-connect option for your Wi-Fi and internet connection for convenience. However, leaving this setting activated means you might accidentally connect to a nefarious network when traveling.

So, before you travel, turn off this feature on your phone, laptop, tablet, and other electronic devices, just to be safe.

<p> While it is important to ensure your internet connection is secure, it’s equally vital that the devices you use to connect to the web are also secure. </p> <p> Before you travel, fully update all electronic devices you plan to take. And if you have downloaded any security software, make sure that gets updated as well. </p> <p>  <a href="https://financebuzz.com/money-moves-after-40?utm_source=msn&utm_medium=feed&synd_slide=4&synd_postid=15545&synd_backlink_title=Grow+Your+%24%24%3A+11+brilliant+ways+to+build+wealth+after+40&synd_backlink_position=4&synd_slug=money-moves-after-40"><b>Grow Your $$:</b> 11 brilliant ways to build wealth after 40</a>  </p>

Update your devices

While it is important to ensure your internet connection is secure, it’s equally vital that the devices you use to connect to the web are also secure.

Before you travel, fully update all electronic devices you plan to take. And if you have downloaded any security software, make sure that gets updated as well.

Grow Your $$: 11 brilliant ways to build wealth after 40

<p> There are a host of clever travel scams that data thieves can use to steal your information. One of the more difficult-to-detect schemes is “juice jacking.” </p> <p>To commit this crime, scammers could potentially load malware on USB port charging stations. You can find these free stations in busy areas such as airports and the lobbies of hotels.  </p> <p> It is important to note that while “juice jacking” is technically possible, the Federal Communications Commission says there are no confirmed instances where it has been used.  </p> <p> Still, it makes sense to play it safe and use your own portable charger to charge devices. </p>

Bring your own chargers

There are a host of clever travel scams that data thieves can use to steal your information. One of the more difficult-to-detect schemes is “juice jacking.”

To commit this crime, scammers could potentially load malware on USB port charging stations. You can find these free stations in busy areas such as airports and the lobbies of hotels.

It is important to note that while “juice jacking” is technically possible, the Federal Communications Commission says there are no confirmed instances where it has been used.

Still, it makes sense to play it safe and use your own portable charger to charge devices.

<p> Odds are good that your email, social media accounts, and many apps have asked you to enable two-factor authentication. Make sure you do so before traveling so you can protect your data. </p> <p> Instead of simply requiring a password, two-factor authentication also asks for a code or PIN. This request is usually sent to another device, such as your phone or computer.  </p> <p> While two-factor authentication doesn’t make your accounts hack-proof, it does provide a bit of extra security.  </p>

Enable two-factor authentication

Odds are good that your email, social media accounts, and many apps have asked you to enable two-factor authentication. Make sure you do so before traveling so you can protect your data.

Instead of simply requiring a password, two-factor authentication also asks for a code or PIN. This request is usually sent to another device, such as your phone or computer.

While two-factor authentication doesn’t make your accounts hack-proof, it does provide a bit of extra security.

<p> Securing your data isn’t just about protecting it from a security breach. Instead, you also want to prepare for such a breach in case it happens despite your best efforts.  </p> <p> Backing up information can help you recover if your data is breached. If you're hacked and have to wipe your devices, it helps to have that information backed up safely at home.  </p> <p> So, transfer a copy of this important data to a USB or external hard drive.  </p> <p>  <a href="https://financebuzz.com/retire-early-quiz?utm_source=msn&utm_medium=feed&synd_slide=7&synd_postid=15545&synd_backlink_title=Retire+Sooner%3A+Take+this+quiz+to+see+if+you+can+retire+early&synd_backlink_position=5&synd_slug=retire-early-quiz"><b>Retire Sooner:</b> Take this quiz to see if you can retire early</a>  </p>

Back up your data

Securing your data isn’t just about protecting it from a security breach. Instead, you also want to prepare for such a breach in case it happens despite your best efforts.

Backing up information can help you recover if your data is breached. If you're hacked and have to wipe your devices, it helps to have that information backed up safely at home.

So, transfer a copy of this important data to a USB or external hard drive.

Retire Sooner: Take this quiz to see if you can retire early

<p> An easy way for hackers to steal your personal information is by stealing your devices. The risk of such a theft looms large when traveling, especially to popular tourist destinations. </p> <p> If your device is stolen, having a tracker or manager can be a tremendous help, as it can help you track where the device is.  </p> <p> Apple has a tracker called Find My, which can track any number of Apple products. Android has Google Find My Device.  </p>

Use device managers

An easy way for hackers to steal your personal information is by stealing your devices. The risk of such a theft looms large when traveling, especially to popular tourist destinations.

If your device is stolen, having a tracker or manager can be a tremendous help, as it can help you track where the device is.

Apple has a tracker called Find My, which can track any number of Apple products. Android has Google Find My Device.

<p> A strong password is one of the best ways to protect your accounts and information. Changing your passwords frequently works even better. </p> <p> So, change passwords before you travel to something especially strong. When you return home, change passwords again. That way, if anyone stole your passwords while you were traveling, they couldn’t get into your accounts.  </p> <p> Also, ensure each account has its unique password rather than sharing a single one among all accounts.  </p>

Change passwords before and after your trip

A strong password is one of the best ways to protect your accounts and information. Changing your passwords frequently works even better.

So, change passwords before you travel to something especially strong. When you return home, change passwords again. That way, if anyone stole your passwords while you were traveling, they couldn’t get into your accounts.

Also, ensure each account has its unique password rather than sharing a single one among all accounts.

<p> While many hackers use advanced, sneaky techniques to steal your data, some hackers use simple tricks like looking over your shoulder and snooping.  </p> <p> Using a VPN can help protect your data, but it won’t stop people from snooping. If you’re using a phone or laptop in public, be careful to shield your screen. </p> <p> Ensure that people aren’t watching you when you enter sensitive information, such as passwords or credit card numbers.  </p> <p>  <a href="https://financebuzz.com/southwest-booking-secrets-55mp?utm_source=msn&utm_medium=feed&synd_slide=10&synd_postid=15545&synd_backlink_title=9+nearly+secret+things+to+do+if+you+fly+Southwest&synd_backlink_position=6&synd_slug=southwest-booking-secrets-55mp">9 nearly secret things to do if you fly Southwest</a>  </p>

Be discrete

While many hackers use advanced, sneaky techniques to steal your data, some hackers use simple tricks like looking over your shoulder and snooping.

Using a VPN can help protect your data, but it won’t stop people from snooping. If you’re using a phone or laptop in public, be careful to shield your screen.

Ensure that people aren’t watching you when you enter sensitive information, such as passwords or credit card numbers.

9 nearly secret things to do if you fly Southwest

<p> If you’re traveling for work or school, consider asking your company or school if it will lend you a work-specific phone or computer so you can protect your personal information. </p> <p> These devices usually have their own encryption and can be wiped clean of data when you get back home and return them to your company’s IT department.  </p>

Ask for a loaner device

If you’re traveling for work or school, consider asking your company or school if it will lend you a work-specific phone or computer so you can protect your personal information.

These devices usually have their own encryption and can be wiped clean of data when you get back home and return them to your company’s IT department.

<p> While it may seem obvious, an important way to protect the data on your devices is by keeping an eye on the devices themselves. Don’t leave your phone or computer unsupervised in public spaces, and be wary of pickpockets in popular tourist spots. </p> <p> Similarly, keep your devices locked in the safety lockbox when leaving your hotel room. That way, no one with access to your room can steal your information. </p>

Keep an eye on your devices

While it may seem obvious, an important way to protect the data on your devices is by keeping an eye on the devices themselves. Don’t leave your phone or computer unsupervised in public spaces, and be wary of pickpockets in popular tourist spots.

Similarly, keep your devices locked in the safety lockbox when leaving your hotel room. That way, no one with access to your room can steal your information.

<p> Most of our information is stored digitally now, including data associated with our bank accounts and credit cards. However, if you can avoid storing it digitally when traveling, you should. </p> <p> For example, make sure your devices do not carry lists of passwords or your Social Security number, driver’s license number, or address. </p> <p>  <a href="https://financebuzz.com/top-travel-credit-cards?utm_source=msn&utm_medium=feed&synd_slide=13&synd_postid=15545&synd_backlink_title=Earn+Points+and+Miles%3A+Find+the+best+travel+credit+card+for+nearly+free+travel&synd_backlink_position=7&synd_slug=top-travel-credit-cards"><b>Earn Points and Miles:</b> Find the best travel credit card for nearly free travel</a>  </p>

Limit what you store digitally

Most of our information is stored digitally now, including data associated with our bank accounts and credit cards. However, if you can avoid storing it digitally when traveling, you should.

For example, make sure your devices do not carry lists of passwords or your Social Security number, driver’s license number, or address.

<p>In today's digital world, a lost passport is no longer the only worry on your travel itinerary. </p><p>Data breaches and identity theft can loom large, casting a shadow over your hard-earned vacation. Don't let your digital security be an afterthought. </p><p>Pack these precautionary tips alongside your sunscreen and <a href="https://financebuzz.com/top-travel-credit-cards?utm_source=msn&utm_medium=feed&synd_slide=14&synd_postid=15545&synd_backlink_title=top+travel+credit+cards&synd_backlink_position=8&synd_slug=top-travel-credit-cards">top travel credit cards</a>, and safeguard your data so you can truly enjoy every moment of your adventure.</p> <p>  <p><b>More from FinanceBuzz:</b></p> <ul> <li><a href="https://financebuzz.com/supplement-income-55mp?utm_source=msn&utm_medium=feed&synd_slide=14&synd_postid=15545&synd_backlink_title=7+things+to+do+if+you%27re+scraping+by+financially.&synd_backlink_position=9&synd_slug=supplement-income-55mp">7 things to do if you're scraping by financially.</a></li> <li><a href="https://www.financebuzz.com/shopper-hacks-Costco-55mp?utm_source=msn&utm_medium=feed&synd_slide=14&synd_postid=15545&synd_backlink_title=6+genius+hacks+Costco+shoppers+should+know.&synd_backlink_position=10&synd_slug=shopper-hacks-Costco-55mp">6 genius hacks Costco shoppers should know.</a></li> <li><a href="https://www.financebuzz.com/diversify-portfolio-fine-art?utm_source=msn&utm_medium=feed&synd_slide=14&synd_postid=15545&synd_backlink_title=See+what+could+happen+if+you+add+fine+art+to+your+investment+portfolio.&synd_backlink_position=11&synd_slug=diversify-portfolio-fine-art">See what could happen if you add fine art to your investment portfolio.</a></li> <li><a href="https://financebuzz.com/extra-newsletter-signup-testimonials-synd?utm_source=msn&utm_medium=feed&synd_slide=14&synd_postid=15545&synd_backlink_title=9+simple+ways+to+make+up+to+an+extra+%24200%2Fday&synd_backlink_position=12&synd_slug=extra-newsletter-signup-testimonials-synd">9 simple ways to make up to an extra $200/day</a></li> </ul>  </p>

Bottom line

In today's digital world, a lost passport is no longer the only worry on your travel itinerary. 

Data breaches and identity theft can loom large, casting a shadow over your hard-earned vacation. Don't let your digital security be an afterthought. 

Pack these precautionary tips alongside your sunscreen and top travel credit cards , and safeguard your data so you can truly enjoy every moment of your adventure.

More from FinanceBuzz:

  • 7 things to do if you're scraping by financially.
  • 6 genius hacks Costco shoppers should know.
  • See what could happen if you add fine art to your investment portfolio.
  • 9 simple ways to make up to an extra $200/day

More for You

Scarlett Johansson demands AI app stop using her likeness in an ad without her permission

Scarlett Johansson says she was 'shocked, angered' when she heard OpenAI's ChatGPT voice that sounded like her

The Most Valuable Hot Wheels Cars on the Market

The Most Valuable Hot Wheels Cars on the Market

Herrera won 10 of his 13 MMA fights

Ex-UFC fighter Herrera dies aged 33

Billionaire Dan Snyder funded a movie about Trump. Now he's reportedly furious it's not flattering.

Billionaire Dan Snyder funded a movie about Trump. Now he's reportedly furious it's not flattering.

Dark, leafy greens boost brain health. Here are 3 tricks to make them tender and delicious without cooking

Kale had its moment. Its leafy cousin has more protein and fewer calories

Chicago Sky forward Angel Reese.

Sky win first game of season, Angel Reese sets WNBA record

happy woman remote worker laptop computer_iStock-1191849990

How Many People Actually Have $1 Million Saved for Retirement?

Legendary singer who helped give The Beatles their start dies aged 86

Legendary singer who helped give The Beatles their start dies aged 86

Pokemon: 10 Best Non-Legendary Poison Types In The Series, Ranked

Pokemon: 10 Best Non-Legendary Poison Types In The Series, Ranked

Kari Lake chances winning senate polls

Kari Lake's Chances at Winning Senate Race, According to Polls

18 Unnecessary Car Features That Will Empty Your Wallet

18 Unnecessary Car Features That Will Empty Your Wallet

20 facts you might not know about 'Barbie'

20 facts you might not know about 'Barbie'

Eminem’s oldest child Hailie Jade gets married to Evan McClintock

Eminem’s oldest child Hailie Jade gets married to Evan McClintock

Bill Maher details his new book on ‘Gutfeld!’

Bill Maher details his new book on ‘Gutfeld!’

Ohio-Class Submarine Transits Suez Canal

North Korea Reacts to 'Nuclear Threat' From US

Tributes pour in as 90s nu-metal drummer dies aged 53

Tributes pour in as 90s nu-metal drummer dies aged 53

A financial planner shares 3 pieces of money advice clients never want to hear

A financial planner shares 3 pieces of money advice clients never want to hear

The New Math of Driving Your Car Till the Wheels Fall Off

The New Math of Driving Your Car Till the Wheels Fall Off

Wonder Woman

25 Unforgettable Female Characters in Fiction: Cherished Heroines

AP19103594147564.jpg

Elvis’ granddaughter fights Graceland foreclosure sale and alleges fraud

Human-sounding AI can plan, help book your travel. But can you trust it?

travel social data

It wasn’t so long ago that travelers planned trips without the internet.

“Back in the day, our parents used to go to these travel agents and really kind of express what they were looking for and what kind of vacation they wanted,” said Saad Saeed, co-founder and CEO of Layla, an AI travel planner whose website launched this year. “Slowly, we kind of acclimatized ourselves to start using these search boxes, clicks, these forms and filters.”

Artificial intelligence-driven tools like Layla can now turn back the clock on that experience, engaging with users almost like humans to customize travel plans with lightning speed plus all the resources of the web. But does AI actually make travel planning easier and can it compare to human expertise? 

Yes and no. Here’s why.

Can AI actually understand us?

It can try. 

“What are you personally looking for in this trip and what do you want out of it?” asked Saeed. “Do you want to reconnect with your partner, for example, or do you want to just feel some adventure and thrill?” 

A human travel agent may ask a series of questions to understand a client’s needs. So can generative AI , which picks up on keywords. Mindtrip, an AI planner launched publicly on May 1, has an actual travel quiz that asks users to rank priorities like “Is your ideal vacation day an exhilarating adventure or a relaxing break?” using sliding scales.

“What we get at the end of that quiz, using the AI, is a really customized description,” explained  Mindtrip Founder and CEO Andy Moss. That then informs what the AI suggests to the traveler. 

Informed suggestions can save users time in narrowing down destinations and experiences, as well as  introduce places users may never have discovered on their own.

AI travel planning is here: How to use it to plan your next vacation and what you should know first

Can AI fully replace humans?

No. Layla may sound human, using conversational phrases like “I've got three cozy nests that won't make your wallet cry.”

“She has a personality. We try to make her funny and so on, where it's really that friend that can get to know you and then recommend you the perfect stuff,” Saeed said.

But part of Layla’s expertise comes from the real-life experiences of some 1,600 travel content creators  the Berlin-based platform has partnered with. Their videos and insights can give users a richer picture of what to expect.

Mindtrip also leans on human expertise, having tapped a limited group of travel influencers for curated content with plans to eventually open it up so anyone can share their travel itineraries and experiences with the public.

Story continues below.

Is AI a threat to privacy?

With all the rapid advancements in AI in just the past year, some users are wary of its safety .

“Data privacy is definitely one of our biggest concerns, and we ensure that none of the personal identifiable information ever reaches basically the model providers. That will all stay with us,” Layla’s Saeed said. “None of their personally identifiable data can ever be basically used to profile them or basically go into any of these systems, which are training these different models.”

Booz Allen Hamilton, the nation’s largest provider of AI to the federal government , focuses heavily on ethical and  secure AI, as well as adhering to the government’s policies on data collection. 

“We collect as little information as we can in order to provide a secure transaction,” said Booz Allen Hamilton Senior Vice President Will Healy, who heads up their recreation work, including Recreaton.gov , the government’s central travel planning site for public lands like national parks. “We don't save your searches. We don't save your credit card data. We're very careful about the data that we store.”

Yoon Kim, an assistant professor in MIT’s Electrical Engineering and Computer Science Department and Computer Science and Artificial Intelligence Laboratory , isn’t too worried about security in the initial brainstorming stages of travel planning with AI.

“I don't see, at this point, how AI-generated advice is spiritually different from travel guide articles that you might read on certain websites,” he said. “Travel planning is one really nice use case of these models, as narrow as it is, because it's a scenario in which you want to be given ideas but you don't actually need to commit to them.” 

What’s next for AI? 

Things could be different, though, if AI is used beyond trip planning. Deloitte sees AI being woven into all parts of travel.

“There is an opportunity for a real engine – I'm going to just use a generic term, engine – that allows you to search and pull it all together and to sort based off of your personal reasons for prioritization and then not stopping at ‘hey give me a list’ or ‘here's what to do,’ but ‘OK, now go create my itinerary, help me book it, track it all the way through that travel process,” said Matt Soderberg, principal, U.S. airlines leader for Deloitte. 

Deloitte’s Facing travel’s future report, released in early April, identifies seven stages where AI can intersect with a trip, from personalized recommendations based on past travel, online purchases and tendencies to day-of issues to a post-travel pulse, where travelers may be asked about their experience and start thinking about future trips. 

“When you solve across all of those, that's going to be the Holy Grail,” Soderberg said. “The difficulty is that doesn't all sit in one place. And so how do you get the right information and the right data to bring all of that together for a single experience for the consumer? And who's going to own that?”

Layla and Mindtrip, among others, already offer booking through partners like Booking.com. “It's all about making things actionable,” Moss said.

But for now, if issues come up mid-trip, AI tools can’t fix them like humans can. Humans still have to get involved.

The Cinderella Castle at Disney World.

13 North American amusement parks growing the most in popularity

Heart-pounding thrill rides, tasty treats, unique attractions—who doesn't love a trip to an amusement park? In fact, our universal (pun intended) love for amusement parks stretches back much further than you might expect: all the way to 1583. That's when the world's oldest amusement park, Bakken, opened in Denmark.

People around the world still flock to amusement parks in the millions today. But when the COVID-19 pandemic struck in 2020, many of the most popular parks in North America temporarily closed their gates or, at the very least, saw significant drops in attendance. Some parks started to recuperate in 2021 as social distancing restrictions eased, but most stayed relatively empty until 2022.

When 2022 did roll around, though, certain parks saw 40-50% spikes in attendance compared to the year prior, while one park's attendance soared by more than 500%. The industry is now firmly back in business, and many amusement parks have implemented new rides, shows, and other exciting experiences to entice visitors to return.

To showcase the parks that saw the biggest increases in annual visitorship between 2021 and 2022, Stacker looked at data from the Themed Entertainment Association and AECOM . Read on to find out which 13 North American amusement parks have grown the most in popularity in recent years.

The Velocicoaster.

#13. Universal Studios Florida

- Location: Orlando, Florida - Attendance growth, 2021-2022: 20% - 2022 attendance: 10.8 million

Though COVID-19 restrictions in Florida were among the most relaxed in the country, Universal Studios Florida still underwent a major dip in visitorship during the pandemic. The park welcomed slightly less than 4.1 million attendees in 2020, compared to the 10.9 million visitors who entered Universal in 2019. But attendance jumped back up in 2022, further bolstered by the 2021 opening of the Jurassic World VelociCoaster elsewhere in the park and a stunt show called the Bourne Stuntacular, which started in 2020.

A water rollercoaster.

#12. Universal's Islands of Adventure

- Location: Orlando, Florida - Attendance growth, 2021-2022: 21% - 2022 attendance: 11.0 million

According to the Themed Entertainment Association and AECOM, Universal's Islands of Adventure accounted for the lion's share of attendance at Universal Orlando Resort in 2022. In fact, visitor numbers at this park alone outstripped three separate Disney World parks: Epcot, Animal Kingdom, and Hollywood Studios. You can find the aforementioned Jurassic World VelociCoaster here, as well as Hagrid's Magical Creatures Motorbike Adventure, a popular ride that opened in 2019.

The Emperor Dive Coaster.

#11. SeaWorld San Diego

- Location: San Diego - Attendance growth, 2021-2022: 25% - 2022 attendance: 3.5 million

Attendance at SeaWorld San Diego in 2022 neared pre-pandemic numbers, which reached 3.7 million in 2019. That could be thanks to the opening of a buzzy new ride, the Emperor Dive Coaster. The tallest and fastest dive coaster in California, the Emperor boasts a breathtaking, 14-story vertical drop. A Mardi Gras event also kicked off for the first time in 2022.

A lion sleeping.

#10. Disney's Animal Kingdom at Walt Disney World

- Location: Lake Buena Vista, Florida - Attendance growth, 2021-2022: 25% - 2022 attendance: 9.0 million

Disney's Animal Kingdom still has a ways to go to reach pre-pandemic attendance (nearly 14 million visitors came to the park in 2019), but the return of several popular tours, like the Up Close with Rhinos experience and the Wild Africa Trek, helped boost visitorship in 2022. Several key, holiday-themed shows also reappeared, including the Merry Menagerie, a wildlife-themed puppet show, and the park's annual Earth Day celebration.

People riding a rollercoaster.

#9. Busch Gardens Tampa Bay

- Location: Tampa, Florida - Attendance growth, 2021-2022: 26% - 2022 attendance: 4.1 million

2022 brought both new events and returning favorites to Busch Gardens Tampa Bay. Celebrations like Cinco de Mayo and Viva La Música took place for the first time, while Mardi Gras, the Food & Wine Festival, and the Real Music concert series returned to the park that year. Busch Gardens also drew crowds to a new ride in March 2022: Iron Gwazi, the tallest hybrid roller coaster in North America.

Notably, Busch Gardens' parent company, SeaWorld, reported increased revenue in 2022 compared to 2019, indicating that while there might be fewer visitors, those visitors are happy to spend more money on their experiences.

Star Wars Park at Hollywood Studios.

#8. Disney's Hollywood Studios at Walt Disney World

- Location: Lake Buena Vista, Florida - Attendance growth, 2021-2022: 27% - 2022 attendance: 10.9 million

Between 2021 and 2022, Disney's Hollywood Studios at Walt Disney World didn't have many new attractions to offer visitors. But that didn't stop over a million more attendees from streaming into the park in 2022, as pandemic restrictions continued to lift and Disney celebrated its 50th anniversary. Disney theme parks did roll out the Genie+ line-skipping service starting in 2021, which could have helped increase attendance at Hollywood Studios and other Walt Disney World attractions.

Epcot Center entrance.

#7. Epcot at Walt Disney World

- Location: Lake Buena Vista, Florida - Attendance growth, 2021-2022: 29% - 2022 attendance: 10.0 million

Several hugely popular events made their regularly scheduled returns to Epcot at Walt Disney World in 2022, including the Festival of the Arts, which features the Disney on Broadway Concert series. In the summer, Epcot also launched The Guardians of the Galaxy Cosmic Rewind, a ride that ranks as one of the world's longest enclosed coasters. Last but not least, the fast-casual restaurant Connections Café and Eatery serving up Starbucks favorites opened its doors for the first time in 2022, a major plus for visitors needing a caffeine boost.

Fireworks over the Cinderella Castle at Disney.

#6. Magic Kingdom Theme Park at Walt Disney World

- Location: Lake Buena Vista, Florida - Attendance growth, 2021-2022: 35% - 2022 attendance: 17.1 million

There's good news and bad news for Magic Kingdom fans. This was the most-visited theme park in the world in 2022, but these visitor stats don't even come close to pre-pandemic numbers, which reached 21 million in 2019. The park also experienced significant construction delays on one of its most anticipated rides, Tron Lightcycle/Run, which was originally slated to open in 2022 but was pushed to 2023.

Mado and Kraken rollercoasters at night.

#5. SeaWorld Orlando

- Location: Orlando, Florida - Attendance growth, 2021-2022: 46% - 2022 attendance: 4.5 million

Just like its West Coast location, SeaWorld Orlando welcomed visitors to a brand-new roller coaster in 2022. The Ice Breaker features four different launches (forward and reverse, in case you were wondering) and the steepest beyond vertical drop in Florida, clocking in at 93 feet tall.

Aside from the Ice Breaker, several key attractions, like the Kraken, were updated and given facelifts. The Electric Ocean summertime event also returned to dazzle audiences with two new shows: stunt show Adrenaline and luminescent percussion concert Electroblast.

People walking through Jupiter's Claim at Universal Studios.

#4. Universal Studios Hollywood

- Location: Universal City, California - Attendance growth, 2021-2022: 53% - 2022 attendance: 8.4 million

Universal Studios Hollywood made some major moves in 2022, namely opening a new film set, Jupiter's Claim, the Western-themed attraction at the center of the horror film "Nope." Guests can explore Jupiter's Claim and other sets aboard the studio tour, which takes visitors behind the scenes of famous Universal releases.

The park's ultrapopular Christmas event, The Awesomest Celebration of the Season, also made a triumphant comeback that year. Festivities took place throughout the Wizarding World of Harry Potter and over in Whoville, where Universal's trademark Grinchmas came to town.

Raya greeting visitors at Disney.

#3. Disney California Adventure Park at Disneyland Resort

- Location: Anaheim, California - Attendance growth, 2021-2022: 81% - 2022 attendance: 9.0 million

Considering all Disneyland Resort parks were closed for nearly four months in 2021 due to COVID-19 restrictions, and Disneyland offered discounted tickets for Southern California residents in 2022, it's no surprise visitor numbers spiked in the latter year. Several long-missed celebrations, including the Lunar New Year and the Food & Wine Festival, also returned to Disney California Adventure Park. Plus, the character of Raya from the 2021 animated film "Raya and the Last Dragon" made her debut here in 2022.

Guests line up to enter Disneyland Park.

#2. Disneyland Park at Disneyland Resort

- Location: Anaheim, California - Attendance growth, 2021-2022: 97% - 2022 attendance: 16.9 million

Attendance at Disneyland Park at Disneyland Resort nearly doubled between 2021 and 2022, likely due to the resort's post-pandemic reopening and those aforementioned discounted tickets. But also in 2022, the park's beloved nighttime shows (known as "spectaculars") finally kicked off again after the long hiatus. Displays like the Main Street Electrical Parade and Disneyland Forever illuminated the park after dark once more, heralding an eagerly anticipated return to normalcy for visitors.

Beautiful floral gardens at Canada's Wonderland.

#1. Canada's Wonderland

- Location: Maple, Ontario - Attendance growth, 2021-2022: 524% - 2022 attendance: 3.8 million

The largest theme park in Canada, Canada's Wonderland had a bumpy year in 2022, with one attraction catching fire and others trapping riders during a storm , but millions still entered the park in 2022, a dramatic increase from the prior year. Canada's Wonderland was especially busy in 2022—new events like the Taste of the Caribbean took place, and the water park attraction Mountain Bay Cliffs opened this year, as did the Lazy Bear Lodge restaurant.

Story editing by Carren Jao. Copy editing by Paris Close. Photo selection by Lacy Kerrick.

Trending Now

Milestones in women's history from the year you were born.

travel social data

Best black and white films of all time

Actors Gloria Swanson and William Holden on a couch in the movie 'Sunset Boulevard.'

50 most meaningful jobs in America

travel social data

Best Law & Order SVU episodes

Actors Mariska Hargitay and Christopher Meloni filming on location for 'Law & Order: SVU' on the streets of Manhattan on Feb. 14, 2011 in New York City.

travel social data

An official website of the United States government

Here’s how you know

travel social data

Official websites use .gov A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS A lock ( Lock A locked padlock ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

CBP Seal, U.S. Customs and Border Protection:  U.S. Department of Homeland Security. Links to CBP.gov homepage

  • National Media Release

CBP Releases April 2024 Monthly Update

WASHINGTON – U.S. Customs and Border Protection (CBP) released operational statistics today for April 2024. CBP monthly reporting can be viewed on CBP’s Stats and Summaries webpage.

“CBP continues to surge resources and personnel to impacted sectors along the border to ensure the safe, swift, and orderly processing of individuals to maximize expedited removals. We have redoubled our efforts, in coordination with partners throughout the hemisphere and around the world, to disrupt the criminal organizations and transportation networks who are putting vulnerable migrants in danger while peddling lies and profiting from them. We have executed the largest surge of removals and disruptive activities against human smuggling networks in the past decade,” said Troy A. Miller, Senior Official Performing the Duties of the Commissioner . “As a result of this increased enforcement, southwest border encounters have not increased, bucking previous trends. We will remain vigilant to continually shifting migration patterns. We are still experiencing challenges along the borders and the nation’s immigration system is not appropriately resourced to handle them, so we continue to call on Congress to take action that would provide our personnel with additional resources and tools.”

CBP continues to work tirelessly to strengthen border security and enforcement efforts, including collaborating with U.S. Immigration and Customs Enforcement (ICE) and U.S. Citizenship and Immigration Services (USCIS) to quickly process noncitizens encountered at the border and remove or return those who do not establish a legal basis to remain in the United States, delivering strengthened consequences promulgated by the Circumvention of Lawful Pathways rule and its associated measures. Since the lifting of Title 42 May 12, 2023 to April 30, 2024, DHS has removed or returned over 720,000 individuals, the vast majority of whom crossed the southwest border, including more than 109,000 individual family members. Total removals and returns since mid-May 2023 exceed removals and returns in every full fiscal year since 2011.

Below are key operational statistics for CBP’s primary mission areas in April 2024. View all CBP statistics online.

Ensuring Border Security and Managing Migration

CBP continues to expeditiously process, remove, and strengthen consequences for individuals who cross our borders unlawfully. Individuals and families without a legal basis to remain in the U.S. are subject to removal pursuant to Title 8 authorities and are subject to a minimum five-year bar on admission as well as potential prosecution if they subsequently re-enter without authorization. No one should believe the lies of smugglers. The fact is that people without a legal basis to remain in the United States will be removed.

The United States is working together with our domestic and foreign partners to jointly disrupt irregular migration across the region, offering safe, orderly, and lawful pathways for intending migrants and taking action against ruthless smugglers who continue to spread falsehoods and show disregard for the safety and well-being of vulnerable migrants.

In April 2024, the U.S. Border Patrol recorded 128,900 encounters between ports of entry along the southwest border. In April, encounters between ports of entry along the southwest border were 6% lower than in March 2024 and 30% lower than April 2023.

CBP continually analyzes and responds to changes in migration patterns, particularly irregular migration outside of lawful pathways and border crossings. We work with our federal and international partners to combat human smuggling. The fact remains: the United States continues to enforce immigration law, and those without a legal basis to remain will be removed. Migrants attempting to enter without authorization are subject to removal under Title 8 authorities.

The U.S. Border Patrol has undertaken significant efforts in recent years to expand capacity to aid and rescue individuals in distress. To prevent the loss of life, CBP initiated a Missing Migrant Program in 2017 that locates noncitizens reported missing, rescues individuals in distress, and reunifies decedents’ remains with their families in the border region. In April, the U.S. Border Patrol conducted 411 rescues, bringing the FY 2024 total to 3,096 rescues.

View more migration statistics and rescues statistics .

CBP One™ App

The CBP One™ mobile application remains a key scheduling tool as part of DHS’s efforts to incentivize noncitizens to use lawful, safe, humane, and orderly pathways and processes. Generally, noncitizens who cross between the ports of entry or who present themselves at a port of entry without making a CBP One™ appointment are subject to the Circumvention of Lawful Pathways rule. This rule presumes asylum ineligibility for those who fail to use lawful processes, with certain exceptions. DHS encourages migrants to utilize lawful processes, rather than taking the dangerous journey to cross unlawfully between the ports of entry, which also carries consequences under Title 8.

The CBP One™ app allows noncitizens throughout central or northern Mexico who lack documents sufficient for admission to the United States to schedule an appointment and remain in place until presenting at a preferred port of entry for their appointment, reducing migrants’ need to crowd into immediate border areas. Use of the CBP One™ app to schedule appointments at ports of entry has increased CBP’s capacity to process migrants more efficiently and orderly while cutting out unscrupulous smugglers who endanger and profit from vulnerable migrants.

In April, CBP processed 41,400 individuals through appointments at ports of entry utilizing advanced information submitted in CBP One™. Since the appointment scheduling function in CBP One™ was introduced in January 2023 through the end of April 2024, more than 591,000 individuals have successfully scheduled appointments to present at ports of entry instead of risking their lives in the hands of smugglers. The top nationalities processed subsequent to arrival for their appointment are Cuban, Haitian, Honduran, Mexican, and Venezuelan.

A percentage of daily available appointments are allocated to the earliest registered CBP One™ profiles, so noncitizens who have been trying to obtain appointments for the longest time are prioritized. CBP is continually monitoring and evaluating the application to ensure its functionality and guard against bad actors.

CHNV Parole Processes

On January 5, 2023, DHS announced processes providing certain Cubans, Haitians, and Nicaraguans who have a supporter in the United States undergo and clear robust security vetting and meet other eligibility criteria authorization to travel to the United States in a safe, orderly, and lawful way. Once they purchase commercial airline tickets for themselves These processes were built on the success of the process for Venezuelans established in October 2022; they are publicly available online, and DHS has been providing regular updates on their use to the public. This is part of the Administration’s strategy to combine expanded lawful pathways with stronger consequences to reduce irregular migration and have kept hundreds of thousands of people from migrating irregularly.Through the end of April 2024, 434,800 Cubans, Haitians, Nicaraguans, and Venezuelans arrived lawfully on commercial flights and were granted parole under these processes. Specifically, 95,500 Cubans, 184,600 Haitians, 83,800 Nicaraguans, and 109,200 Venezuelans were vetted and authorized for travel; and 91,100 Cubans, 166,700 Haitians, 75,700 Nicaraguans, and 101,200 Venezuelans arrived lawfully and were granted parole.

Safeguarding Communities by Interdicting Narcotics and Dangerous Drugs

As the largest law enforcement agency in the United States, CBP is uniquely positioned to detect, identify, and seize illicit drugs before they enter our communities. CBP’s combination of interdiction and intelligence capabilities, complemented by its border search authorities, scientific services, non-intrusive inspection equipment, and canine detection teams, places it at the forefront of the U.S. government’s efforts to combat illicit fentanyl and other dangerous drugs.

In April, CBP also announced an expanded, multi-agency effort to target transnational criminals funneling fentanyl from Mexico into American communities. Operation Plaza Spike targets the cartels that facilitate the flow of deadly fentanyl, as well as its analogs, precursors, and tools to make the drugs. The operation is designed to disrupt operations in the “plazas,” cartel territories located directly south of the United States that are natural logistical chokepoints within the cartels’ operations. This is the next phase in CBP’s Strategy to Combat Fentanyl and Other Synthetic Drugs , a whole-of-government and international effort to anticipate, identify, mitigate, and disrupt illicit synthetic drug producers, suppliers, and traffickers.

That strategy also includes conducting operations, including Operation Apollo, that target the smuggling of illicit fentanyl and other dangerous drugs. First implemented in southern California in October 2023, and recently expanded into Arizona, Operation Apollo utilizes local field assets augmented by federal, state, local, tribal, and territorial partners to target drug traffickers’ supply chains in select locations based on ongoing investigations, intelligence collection, and drug seizure data. Operation Apollo targets items required in the production of illicit fentanyl, including precursor chemicals, pill presses and parts, movement of finished product, and illicit proceeds.

Nationwide in April, cocaine seizures increased by 95% compared to March. To date in FY 2024 through the end of April, CBP has seized over 11,400 pounds of fentanyl. CBP has caught more fentanyl nationwide between the start of fiscal year 2023 through April 30, 2024 than in the previous five fiscal years combined, and we continue to optimize our intelligence and field operations to stop these deadly substances from reaching American communities.

Additional CBP drug seizure statistics can be found on the Drug Seizure Statistics webpage .

Facilitating Lawful Trade and Travel

As international travel continues to increase, CBP is leveraging technology to streamline efficiency and increase security at air and land ports of entry. Travelers are encouraged to utilize CBP’s mobile apps to enhance their travel experience, including the Global Entry Mobile Application and Mobile Passport Control , as well as new Global Entry Touchless Portals at nearly all international airports across the United States, which protect passenger privacy and expedite arrival processing by eliminating paper receipts.

Commercial trucks processed at ports of entry increased 15% from April 2023 to April 2024. Travelers arriving by air into the United States increased 8% in the same period; passenger vehicles processed at ports of entry increased 3% ; and pedestrians arriving by land at ports of entry increased 2% over the same period.

CBP works diligently with the trade community and port operators to ensure that merchandise is cleared as efficiently as possible and to strengthen international supply chains and improve border security. In April 2024, CBP processed more than 3.2 million entry summaries valued at more than $289 billion, identifying estimated duties of nearly $6.4 billion to be collected by the U.S. government. In April, trade via the ocean environment accounted for 39.66% of the total import value, followed by air, truck, and rail.

View more travel statistics , and trade statistics .

Protecting Consumers, Eradicating Forced Labor from Supply Chains, and Promoting Economic Security

CBP continues to lead U.S. government efforts to eliminate goods from the supply chain made with forced labor from the Xinjiang Uyghur Autonomous Region of China. In April, CBP stopped 392 shipments valued at more than $184 million for further examination based on the suspected use of forced labor.

Intellectual property rights violations continue to put America’s innovation economy at risk. Counterfeit and pirated goods threaten the competitiveness of U.S. businesses, the livelihoods of American workers, and the health and safety of consumers.

Consumers are encouraged to be alert to the dangers of counterfeit goods especially when shopping online as they support criminal activity, hurt American businesses, and often have materials or ingredients that can pose serious health and safety risks. Every year CBP seizes millions of counterfeit products worth billions of dollars had they been genuine. In April, CBP seized 1,736 shipments that contained counterfeit goods valued at more than $235 million . More information about CBP’s intellectual property rights enforcement is available at www.cbp.gov/trade .

CBP completed 20 audits in April that identified $13 million in duties and fees owed to the U.S. government, stemming from goods that had been improperly declared in accordance with U.S. trade laws and customs regulations. CBP collected over $5.7 million of this identified revenue and from previous fiscal years’ assignments.

CBP is on the frontline of textiles and trade agreements enforcement, combating textile imports that are not compliant with U.S. trade laws. Protecting the domestic textile industry and American consumers is vital to U.S. national security, health care, and economic priorities. Toward this end, CBP is intensifying its targeting and enforcement efforts to increase and expedite the prosecution of illegal customs practices. CBP’s efforts include de minimis compliance, forced labor enforcement, cargo compliance, regulatory audits, and public awareness. This month DHS announced an enhanced strategy to combat illicit trade and level the playing field for the American textile industry, which accounts for over 500,000 U.S. jobs and is critical for our national security. The plan details the actions CBP and Homeland Security Investigations will take to hold perpetrators accountable for customs violations and safeguard the American textile industry.

View more UFLPA enforcement statistics , and intellectual property rights enforcement statistics .

Defending our Nation’s Agricultural System

Through targeting, detection, and interception, CBP agriculture specialists work to prevent threats from entering the United States.

CBP issued 7,139 emergency action notifications for restricted and prohibited plant and animal products entering the United States in April 2024. CBP conducted 101,416 positive passenger inspections and issued 823 civil penalties and/or violations to the traveling public for failing to declare prohibited agriculture items.

View more agricultural enforcement statistics .

U.S. Customs and Border Protection (CBP) is America's frontline: the nation's largest law enforcement organization and the world's first unified border management agency. The 65,000+ men and women of CBP protect America on the ground, in the air, and on the seas. We facilitate safe, lawful travel and trade and ensure our country's economic prosperity. We enhance the nation's security through innovation, intelligence, collaboration, and trust.

  • Topics ›
  • COVID-19 travel passports ›

Travel becomes more important post-pandemic

Sponsored post by booking.com.

In a 2022 survey conducted by YouGov, 35% of travellers stated that travel has become more important to them since the pandemic and more than two thirds were planning a trip in the next 12 months. Of those respondents planning a trip, around twice as many were planning to travel domestically than those going international.

Description

This infographic shows global consumers' travel attitudes or plans as of the end of 2022.

Can I integrate infographics into my blog or website?

Yes, Statista allows the easy integration of many infographics on other websites. Simply copy the HTML code that is shown for the relevant statistic in order to integrate it. Our standard is 660 pixels, but you can customize how the statistic is displayed to suit your site by setting the width and the display size. Please note that the code must be integrated into the HTML code (not only the text) for WordPress pages and other CMS sites.

Infographic: Travel becomes more important post-pandemic | Statista

Infographic Newsletter

Statista offers daily infographics about trending topics, covering: Economy & Finance , Politics & Society , Tech & Media , Health & Environment , Consumer , Sports and many more.

Related Infographics

Sponsored post by booking.com, sustainable tourism: collective responsibility of providers, travelers, and goverenments, a widening gap: smes struggling to keep up with large enterprises, european hoteliers’ biggest opportunities for 2024, live music industry shatters records in post-covid comeback, china smartphones, apple's china business stays volatile as market calms, 5% of americans are digital nomads, public libraries, u.s. and canadian libraries still struggle with attendance, offline hotel bookings defy digitalization, the evolution of air travel, chinese car exports, chinese car exports continue to soar as ev share grows, social media shapes travel experiences.

  • Who may use the "Chart of the Day"? The Statista "Chart of the Day", made available under the Creative Commons License CC BY-ND 3.0, may be used and displayed without charge by all commercial and non-commercial websites. Use is, however, only permitted with proper attribution to Statista. When publishing one of these graphics, please include a backlink to the respective infographic URL. More Information
  • Which topics are covered by the "Chart of the Day"? The Statista "Chart of the Day" currently focuses on two sectors: "Media and Technology", updated daily and featuring the latest statistics from the media, internet, telecommunications and consumer electronics industries; and "Economy and Society", which current data from the United States and around the world relating to economic and political issues as well as sports and entertainment.
  • Does Statista also create infographics in a customized design? For individual content and infographics in your Corporate Design, please visit our agency website www.statista.design

Any more questions?

Get in touch with us quickly and easily. we are happy to help.

Feel free to contact us anytime using our contact form or visit our FAQ page .

Statista Content & Design

Need infographics, animated videos, presentations, data research or social media charts?

More Information

The Statista Infographic Newsletter

Receive a new up-to-date issue every day for free.

  • Our infographics team prepares current information in a clear and understandable format
  • Relevant facts covering media, economy, e-commerce, and FMCG topics
  • Use our newsletter overview to manage the topics that you have subscribed to

travel social data

Court Appears Ready to Clear Cruise Lines in Cuba Travel Suit

By Peter Hayes

Peter Hayes

The Eleventh Circuit Friday seemed open to absolving four cruise lines of $440 million in liability for committing “trafficking acts” by carrying passengers to Havana and using port facilities that had been confiscated by Fidel Castro’s government in 1960.

Plaintiff Havana Docks Corp. held a 99-year leasehold interest to operate the Havana harbor terminal, which would have expired in 2004. In oral arguments Friday, a three-judge panel of the US Court of Appeals for the Eleventh Circuit suggested that HDC’s rights had expired before the cruise lines began Cuban trips.

After the Obama administration loosened travel restrictions with Cuba in ...

Learn more about Bloomberg Law or Log In to keep reading:

Learn about bloomberg law.

AI-powered legal analytics, workflow tools and premium legal & business news.

Already a subscriber?

Log in to keep reading or access research tools.

A woman standing in a server room holding a laptop connected to a series of tall, black servers cabinets.

Published: 5 April 2024 Contributors: Tim Mucci, Cole Stryker

Big data analytics refers to the systematic processing and analysis of large amounts of data and complex data sets, known as big data, to extract valuable insights. Big data analytics allows for the uncovering of trends, patterns and correlations in large amounts of raw data to help analysts make data-informed decisions. This process allows organizations to leverage the exponentially growing data generated from diverse sources, including internet-of-things (IoT) sensors, social media, financial transactions and smart devices to derive actionable intelligence through advanced analytic techniques.

In the early 2000s, advances in software and hardware capabilities made it possible for organizations to collect and handle large amounts of unstructured data. With this explosion of useful data, open-source communities developed big data frameworks to store and process this data. These frameworks are used for distributed storage and processing of large data sets across a network of computers. Along with additional tools and libraries, big data frameworks can be used for:

  • Predictive modeling by incorporating artificial intelligence (AI) and statistical algorithms
  • Statistical analysis for in-depth data exploration and to uncover hidden patterns
  • What-if analysis to simulate different scenarios and explore potential outcomes
  • Processing diverse data sets, including structured, semi-structured and unstructured data from various sources.

Four main data analysis methods  – descriptive, diagnostic, predictive and prescriptive  – are used to uncover insights and patterns within an organization's data. These methods facilitate a deeper understanding of market trends, customer preferences and other important business metrics.

IBM named a Leader in the 2024 Gartner® Magic Quadrant™ for Augmented Data Quality Solutions.

Structured vs unstructured data

What is data management?

The main difference between big data analytics and traditional data analytics is the type of data handled and the tools used to analyze it. Traditional analytics deals with structured data, typically stored in relational databases . This type of database helps ensure that data is well-organized and easy for a computer to understand. Traditional data analytics relies on statistical methods and tools like structured query language (SQL) for querying databases.

Big data analytics involves massive amounts of data in various formats, including structured, semi-structured and unstructured data. The complexity of this data requires more sophisticated analysis techniques. Big data analytics employs advanced techniques like machine learning and data mining to extract information from complex data sets. It often requires distributed processing systems like Hadoop to manage the sheer volume of data.

These are the four methods of data analysis at work within big data:

The "what happened" stage of data analysis. Here, the focus is on summarizing and describing past data to understand its basic characteristics.

The “why it happened” stage. By delving deep into the data, diagnostic analysis identifies the root patterns and trends observed in descriptive analytics.

The “what will happen” stage. It uses historical data, statistical modeling and machine learning to forecast trends.

Describes the “what to do” stage, which goes beyond prediction to provide recommendations for optimizing future actions based on insights derived from all previous.

The following dimensions highlight the core challenges and opportunities inherent in big data analytics.

The sheer volume of data generated today, from social media feeds, IoT devices, transaction records and more, presents a significant challenge. Traditional data storage and processing solutions are often inadequate to handle this scale efficiently. Big data technologies and cloud-based storage solutions enable organizations to store and manage these vast data sets cost-effectively, protecting valuable data from being discarded due to storage limitations.

Data is being produced at unprecedented speeds, from real-time social media updates to high-frequency stock trading records. The velocity at which data flows into organizations requires robust processing capabilities to capture, process and deliver accurate analysis in near real-time. Stream processing frameworks and in-memory data processing are designed to handle these rapid data streams and balance supply with demand.

Today's data comes in many formats, from structured to numeric data in traditional databases to unstructured text, video and images from diverse sources like social media and video surveillance. This variety demans flexible data management systems to handle and integrate disparate data types for comprehensive analysis. NoSQL databases , data lakes and schema -on-read technologies provide the necessary flexibility to accommodate the diverse nature of big data.

Data reliability and accuracy are critical, as decisions based on inaccurate or incomplete data can lead to negative outcomes. Veracity refers to the data's trustworthiness, encompassing data quality, noise and anomaly detection issues. Techniques and tools for data cleaning, validation and verification are integral to ensuring the integrity of big data, enabling organizations to make better decisions based on reliable information.

Big data analytics aims to extract actionable insights that offer tangible value. This involves turning vast data sets into meaningful information that can inform strategic decisions, uncover new opportunities and drive innovation. Advanced analytics, machine learning and AI are key to unlocking the value contained within big data, transforming raw data into strategic assets.

Data professionals, analysts, scientists and statisticians prepare and process data in a data lakehouse, which combines the performance of a data lakehouse with the flexibility of a data lake to clean data and ensure its quality. The process of turning raw data into valuable insights encompasses several key stages:

  • Collect data: The first step involves gathering data, which can be a mix of structured and unstructured forms from myriad sources like cloud, mobile applications and IoT sensors. This step is where organizations adapt their data collection strategies and integrate data from varied sources into central repositories like a data lake, which can automatically assign metadata for better manageability and accessibility.
  • Process data: After being collected, data must be systematically organized, extracted, transformed and then loaded into a storage system to ensure accurate analytical outcomes. Processing involves converting raw data into a format that is usable for analysis, which might involve aggregating data from different sources, converting data types or organizing data into structure formats. Given the exponential growth of available data, this stage can be challenging. Processing strategies may vary between batch processing, which handles large data volumes over extended periods and stream processing, which deals with smaller real-time data batches.
  • Clean data: Regardless of size, data must be cleaned to ensure quality and relevance. Cleaning data involves formatting it correctly, removing duplicates and eliminating irrelevant entries. Clean data prevents the corruption of output and safeguard’s reliability and accuracy.
  • Analyze data: Advanced analytics, such as data mining, predictive analytics, machine learning and deep learning, are employed to sift through the processed and cleaned data. These methods allow users to discover patterns, relationships and trends within the data, providing a solid foundation for informed decision-making.

Under the Analyze umbrella, there are potentially many technologies at work, including data mining, which is used to identify patterns and relationships within large data sets; predictive analytics, which forecasts future trends and opportunities; and deep learning , which mimics human learning patterns to uncover more abstract ideas.

Deep learning uses an artificial neural network with multiple layers to model complex patterns in data. Unlike traditional machine learning algorithms, deep learning learns from images, sound and text without manual help. For big data analytics, this powerful capability means the volume and complexity of data is not an issue.

Natural language processing (NLP) models allow machines to understand, interpret and generate human language. Within big data analytics, NLP extracts insights from massive unstructured text data generated across an organization and beyond.

Structured Data

Structured data refers to highly organized information that is easily searchable and typically stored in relational databases or spreadsheets. It adheres to a rigid schema, meaning each data element is clearly defined and accessible in a fixed field within a record or file. Examples of structured data include:

  • Customer names and addresses in a customer relationship management (CRM) system
  • Transactional data in financial records, such as sales figures and account balances
  • Employee data in human resources databases, including job titles and salaries

Structured data's main advantage is its simplicity for entry, search and analysis, often using straightforward database queries like SQL. However, the rapidly expanding universe of big data means that structured data represents a relatively small portion of the total data available to organizations.

Unstructured Data

Unstructured data lacks a pre-defined data model, making it more difficult to collect, process and analyze. It comprises the majority of data generated today, and includes formats such as:

  • Textual content from documents, emails and social media posts
  • Multimedia content, including images, audio files and videos
  • Data from IoT devices, which can include a mix of sensor data, log files and time-series data

The primary challenge with unstructured data is its complexity and lack of uniformity, requiring more sophisticated methods for indexing, searching and analyzing. NLP, machine learning and advanced analytics platforms are often employed to extract meaningful insights from unstructured data.

Semi-structured data

Semi-structured data occupies the middle ground between structured and unstructured data. While it does not reside in a relational database, it contains tags or other markers to separate semantic elements and enforce hierarchies of records and fields within the data. Examples include:

  • JSON (JavaScript Object Notation) and XML (eXtensible Markup Language) files, which are commonly used for web data interchange
  • Email, where the data has a standardized format (e.g., headers, subject, body) but the content within each section is unstructured
  • NoSQL databases, can store and manage semi-structured data more efficiently than traditional relational databases

Semi-structured data is more flexible than structured data but easier to analyze than unstructured data, providing a balance that is particularly useful in web applications and data integration tasks.

Ensuring data quality and integrity, integrating disparate data sources, protecting data privacy and security and finding the right talent to analyze and interpret data can present challenges to organizations looking to leverage their extensive data volumes. What follows are the benefits organizations can realize once they see success with big data analytics:

Real-time intelligence

One of the standout advantages of big data analytics is the capacity to provide real-time intelligence. Organizations can analyze vast amounts of data as it is generated from myriad sources and in various formats. Real-time insight allows businesses to make quick decisions, respond to market changes instantaneously and identify and act on opportunities as they arise.

Better-informed decisions

With big data analytics, organizations can uncover previously hidden trends, patterns and correlations. A deeper understanding equips leaders and decision-makers with the information needed to strategize effectively, enhancing business decision-making in supply chain management, e-commerce, operations and overall strategic direction.  

Cost savings

Big data analytics drives cost savings by identifying business process efficiencies and optimizations. Organizations can pinpoint wasteful expenditures by analyzing large datasets, streamlining operations and enhancing productivity. Moreover, predictive analytics can forecast future trends, allowing companies to allocate resources more efficiently and avoid costly missteps.

Better customer engagement

Understanding customer needs, behaviors and sentiments is crucial for successful engagement and big data analytics provides the tools to achieve this understanding. Companies gain insights into consumer preferences and tailor their marketing strategies by analyzing customer data.

Optimized risk management strategies

Big data analytics enhances an organization's ability to manage risk by providing the tools to identify, assess and address threats in real time. Predictive analytics can foresee potential dangers before they materialize, allowing companies to devise preemptive strategies.

As organizations across industries seek to leverage data to drive decision-making, improve operational efficiencies and enhance customer experiences, the demand for skilled professionals in big data analytics has surged. Here are some prominent career paths that utilize big data analytics:

Data scientist

Data scientists analyze complex digital data to assist businesses in making decisions. Using their data science training and advanced analytics technologies, including machine learning and predictive modeling, they uncover hidden insights in data.

Data analyst

Data analysts turn data into information and information into insights. They use statistical techniques to analyze and extract meaningful trends from data sets, often to inform business strategy and decisions.

Data engineer

Data engineers prepare, process and manage big data infrastructure and tools. They also develop, maintain, test and evaluate data solutions within organizations, often working with massive datasets to assist in analytics projects.

Machine learning engineer

Machine learning engineers focus on designing and implementing machine learning applications. They develop sophisticated algorithms that learn from and make predictions on data.

Business intelligence analyst

Business intelligence (BI) analysts help businesses make data-driven decisions by analyzing data to produce actionable insights. They often use BI tools to convert data into easy-to-understand reports and visualizations for business stakeholders.

Data visualization specialist

These specialists focus on the visual representation of data. They create data visualizations that help end users understand the significance of data by placing it in a visual context.

Data architect

Data architects design, create, deploy and manage an organization's data architecture. They define how data is stored, consumed, integrated and managed by different data entities and IT systems.

IBM and Cloudera have partnered to create an industry-leading, enterprise-grade big data framework distribution plus a variety of cloud services and products — all designed to achieve faster analytics at scale.

IBM Db2 Database on IBM Cloud Pak for Data combines a proven, AI-infused, enterprise-ready data management system with an integrated data and AI platform built on the security-rich, scalable Red Hat OpenShift foundation.

IBM Big Replicate is an enterprise-class data replication software platform that keeps data consistent in a distributed environment, on-premises and in the hybrid cloud, including SQL and NoSQL databases.

A data warehouse is a system that aggregates data from different sources into a single, central, consistent data store to support data analysis, data mining, artificial intelligence and machine learning.

Business intelligence gives organizations the ability to get answers they can understand. Instead of using best guesses, they can base decisions on what their business data is telling them — whether it relates to production, supply chain, customers or market trends.

Cloud computing is the on-demand access of physical or virtual servers, data storage, networking capabilities, application development tools, software, AI analytic tools and more—over the internet with pay-per-use pricing. The cloud computing model offers customers flexibility and scalability compared to traditional infrastructure.

Purpose-built data-driven architecture helps support business intelligence across the organization. IBM analytics solutions allow organizations to simplify raw data access, provide end-to-end data management and empower business users with AI-driven self-service analytics to predict outcomes.

  • Skip to main content
  • Keyboard shortcuts for audio player

The huge solar storm is keeping power grid and satellite operators on edge

Geoff Brumfiel, photographed for NPR, 17 January 2019, in Washington DC.

Geoff Brumfiel

Willem Marx

travel social data

NASA's Solar Dynamics Observatory captured this image of solar flares early Saturday afternoon. The National Oceanic and Atmospheric Administration says there have been measurable effects and impacts from the geomagnetic storm. Solar Dynamics Observatory hide caption

NASA's Solar Dynamics Observatory captured this image of solar flares early Saturday afternoon. The National Oceanic and Atmospheric Administration says there have been measurable effects and impacts from the geomagnetic storm.

Planet Earth is getting rocked by the biggest solar storm in decades – and the potential effects have those people in charge of power grids, communications systems and satellites on edge.

The National Oceanic and Atmospheric Administration says there have been measurable effects and impacts from the geomagnetic storm that has been visible as aurora across vast swathes of the Northern Hemisphere. So far though, NOAA has seen no reports of major damage.

Photos: See the Northern lights from rare solar storm

The Picture Show

Photos: see the northern lights from rare, solar storm.

There has been some degradation and loss to communication systems that rely on high-frequency radio waves, NOAA told NPR, as well as some preliminary indications of irregularities in power systems.

"Simply put, the power grid operators have been busy since yesterday working to keep proper, regulated current flowing without disruption," said Shawn Dahl, service coordinator for the Boulder, Co.-based Space Weather Prediction Center at NOAA.

NOAA Issues First Severe Geomagnetic Storm Watch Since 2005

Short Wave

  • LISTEN & FOLLOW
  • Apple Podcasts
  • Google Podcasts
  • Amazon Music
  • Amazon Alexa

Your support helps make our show possible and unlocks access to our sponsor-free feed.

"Satellite operators are also busy monitoring spacecraft health due to the S1-S2 storm taking place along with the severe-extreme geomagnetic storm that continues even now," Dahl added, saying some GPS systems have struggled to lock locations and offered incorrect positions.

NOAA's GOES-16 satellite captured a flare erupting occurred around 2 p.m. EDT on May 9, 2024.

As NOAA had warned late Friday, the Earth has been experiencing a G5, or "Extreme," geomagnetic storm . It's the first G5 storm to hit the planet since 2003, when a similar event temporarily knocked out power in part of Sweden and damaged electrical transformers in South Africa.

The NOAA center predicted that this current storm could induce auroras visible as far south as Northern California and Alabama.

Extreme (G5) geomagnetic conditions have been observed! pic.twitter.com/qLsC8GbWus — NOAA Space Weather Prediction Center (@NWSSWPC) May 10, 2024

Around the world on social media, posters put up photos of bright auroras visible in Russia , Scandinavia , the United Kingdom and continental Europe . Some reported seeing the aurora as far south as Mallorca, Spain .

The source of the solar storm is a cluster of sunspots on the sun's surface that is 17 times the diameter of the Earth. The spots are filled with tangled magnetic fields that can act as slingshots, throwing huge quantities of charged particles towards our planet. These events, known as coronal mass ejections, become more common during the peak of the Sun's 11-year solar cycle.

A powerful solar storm is bringing northern lights to unusual places

Usually, they miss the Earth, but this time, NOAA says several have headed directly toward our planet, and the agency predicted that several waves of flares will continue to slam into the Earth over the next few days.

While the storm has proven to be large, predicting the effects from such incidents can be difficult, Dahl said.

Shocking problems

The most disruptive solar storm ever recorded came in 1859. Known as the "Carrington Event," it generated shimmering auroras that were visible as far south as Mexico and Hawaii. It also fried telegraph systems throughout Europe and North America.

Stronger activity on the sun could bring more displays of the northern lights in 2024

Stronger activity on the sun could bring more displays of the northern lights in 2024

While this geomagnetic storm will not be as strong, the world has grown more reliant on electronics and electrical systems. Depending on the orientation of the storm's magnetic field, it could induce unexpected electrical currents in long-distance power lines — those currents could cause safety systems to flip, triggering temporary power outages in some areas.

my cat just experienced the aurora borealis, one of the world's most radiant natural phenomena... and she doesn't care pic.twitter.com/Ee74FpWHFm — PJ (@kickthepj) May 10, 2024

The storm is also likely to disrupt the ionosphere, a section of Earth's atmosphere filled with charged particles. Some long-distance radio transmissions use the ionosphere to "bounce" signals around the globe, and those signals will likely be disrupted. The particles may also refract and otherwise scramble signals from the global positioning system, according to Rob Steenburgh, a space scientist with NOAA. Those effects can linger for a few days after the storm.

Like Dahl, Steenburgh said it's unclear just how bad the disruptions will be. While we are more dependent than ever on GPS, there are also more satellites in orbit. Moreover, the anomalies from the storm are constantly shifting through the ionosphere like ripples in a pool. "Outages, with any luck, should not be prolonged," Steenburgh said.

What Causes The Northern Lights? Scientists Finally Know For Sure

What Causes The Northern Lights? Scientists Finally Know For Sure

The radiation from the storm could have other undesirable effects. At high altitudes, it could damage satellites, while at low altitudes, it's likely to increase atmospheric drag, causing some satellites to sink toward the Earth.

The changes to orbits wreak havoc, warns Tuija Pulkkinen, chair of the department of climate and space sciences at the University of Michigan. Since the last solar maximum, companies such as SpaceX have launched thousands of satellites into low Earth orbit. Those satellites will now see their orbits unexpectedly changed.

"There's a lot of companies that haven't seen these kind of space weather effects before," she says.

The International Space Station lies within Earth's magnetosphere, so its astronauts should be mostly protected, Steenburgh says.

In a statement, NASA said that astronauts would not take additional measures to protect themselves. "NASA completed a thorough analysis of recent space weather activity and determined it posed no risk to the crew aboard the International Space Station and no additional precautionary measures are needed," the agency said late Friday.

travel social data

People visit St Mary's lighthouse in Whitley Bay to see the aurora borealis on Friday in Whitley Bay, England. Ian Forsyth/Getty Images hide caption

People visit St Mary's lighthouse in Whitley Bay to see the aurora borealis on Friday in Whitley Bay, England.

While this storm will undoubtedly keep satellite operators and utilities busy over the next few days, individuals don't really need to do much to get ready.

"As far as what the general public should be doing, hopefully they're not having to do anything," Dahl said. "Weather permitting, they may be visible again tonight." He advised that the largest problem could be a brief blackout, so keeping some flashlights and a radio handy might prove helpful.

I took these photos near Ranfurly in Central Otago, New Zealand. Anyone can use them please spread far and wide. :-) https://t.co/NUWpLiqY2S — Dr Andrew Dickson reform/ACC (@AndrewDickson13) May 10, 2024

And don't forget to go outside and look up, adds Steenburgh. This event's aurora is visible much further south than usual.

A faint aurora can be detected by a modern cell phone camera, he adds, so even if you can't see it with your eyes, try taking a photo of the sky.

The aurora "is really the gift from space weather," he says.

  • space weather
  • solar flares
  • solar storm

IMAGES

  1. 35+ Stats on How Social Media Affects Travel [2023]

    travel social data

  2. Infographic: How travellers traverse through social media while

    travel social data

  3. Infographic: How travellers traverse through social media while

    travel social data

  4. Destination Marketing: A Guide to Social Media

    travel social data

  5. Social media: the holy grail of post-Covid travel distribution

    travel social data

  6. We asked, you answered: Travel social media apps and tour tips| EF Tours

    travel social data

COMMENTS

  1. Digitalization of the travel industry

    Premium Statistic Travel and tourism comments on social media worldwide 2019-2023, by product type Premium Statistic Digital media importance among DMOs worldwide September 2023

  2. Data Insights

    The U.S. Travel Insights Dashboard is the most comprehensive and centralized source for high-frequency intelligence on the U.S. travel industry and broader economy. The dashboard is updated the last week of every month. Member log-in required. U.S. Travel's Economic Impact Map tells the story of travel's economic impact by state and ...

  3. The U.S. Travel Insights Dashboard

    INTERACTIVE TRAVEL DATA April 04, 2024. U.S. Travel members have access to the exclusive U.S. Travel Insights Dashboard, the most comprehensive and centralized source for high-frequency intelligence on the U.S. travel industry and the broader economy. The platform, powered by Tourism Economics, is supported by more than 20 data partners and ...

  4. How social is revealing a new world for the travel industry

    I dove into branded and non-branded social travel content from both Twitter and Instagram, searching for travel influencers, airline and hotel brands, and everyday travelers sharing details about their trips. Three main patterns emerged from the data. 1. Travel influencers are changing the content game

  5. Millennial travel in the U.S.

    Directly accessible data for 170 industries from 150+ countries and over 1 Mio. facts. ... Opinion on social media presence of online travel brands in the U.S. 2021, by age.

  6. The UN Tourism Data Dashboard

    International Tourism and COVID-19. Export revenues from international tourism dropped 62% in 2020 and 59% in 2021, versus 2019 (real terms) and then rebounded in 2022, remaining 34% below pre-pandemic levels. The total loss in export revenues from tourism amounts to USD 2.6 trillion for that three-year period. Go to Dashboard.

  7. Research & Insights Hub

    We also produce reports on the environmental and social impacts of Travel & Tourism, as well as thought leadership reports on themes such as diversity, equity, inclusion and belonging, and retail tourism. ... Discover the total environmental and social footprint of the Travel & Tourism sector in data-rich, four-page factsheets for 185 economies ...

  8. The roles of social media in tourists' choices of travel components

    The influence of social media on travel decision-making has attracted much attention from tourism scholars. A recent literature analysis has suggested that most of such studies have focused on the impact of social media on behavioural intention with very limited studies on actual behaviour (Leung et al., 2019).Furthermore, relatively little insight has been put on the roles of social media in ...

  9. PDF Travel Social Media Industry Report

    Instagram and Twitter in the Travel business remains steady. Since May 2023, there has been recorded a 97.39% decrease in TikTok interactions. Median posts interactions across all platforms Date Range: 1 Jan 2023 - 30 Jun 2023 Sample: Socialinsider—worldwide data for the Travel industry

  10. Influencers and Social Commerce in Travel

    The confluence of these forces creates a catalyst to engage directly with consumers through social commerce - a rapidly growing sector where travel currently lags. According to one projection, U.S. social commerce will more than double from $37 billion in 2021 to $80 billion in 2025, growing its share of total e-commerce sales to 5.2%.

  11. Travel analytics: Understanding how destination choice and business

    One key contribution of this research is to investigate the methods for analyzing and visualizing geo-tagged social media data to understand destination choice and business clusters. We further discuss how social media data can be used to supplement travel survey data to explore human-environment relationships.

  12. Feasibility of estimating travel demand using geolocations of social

    Travel demand estimation, as represented by an origin-destination (OD) matrix, is essential for urban planning and management. Compared to data typically used in travel demand estimation, the key strengths of social media data are that they are low-cost, abundant, available in real-time, and free of geographical partition. However, the data also have significant limitations: population and ...

  13. 98 Statistics on Travel

    01 General Traveling Statistics. Travel and tourism contribute roughly $5.81 trillion to the global economy. Statista. 1 in 20 jobs depend on the travel industry. U.S. Travel Association. September 2022 saw an increase of 6% in travel spending compared to pre-pandemic levels in 2019.

  14. The Latest Travel Data (2024-03-04)| U.S. Travel Association

    The Latest Travel Data. MONTHLY INSIGHTS March 04, 2024. U.S. Travel has temporarily paused our monthly data newsletter, however, the latest travel data is still available via the U.S. Travel Insights Dashboard. This dashboard is updated each month (member login required). The U.S. Travel Insights Dashboard, developed in collaboration with ...

  15. Data Article Dataset for understanding why people share their travel

    The data presented in this article relates to the individual intrinsic and extrinsic motivations to share travel experience in social media. The 381 records were gathered in Portugal using an online survey. A statistical analysis of the data was carried out using partial least squares (PLS).

  16. Technological Forecasting and Social Change

    The data were collected using three travel social networks, namely Tripadvisor, Minube, and Travel365, with each directed towards capturing focused information more efficiently than others (i.e. Skyscanner and Booking were removed for technical remarks).

  17. Travel Sector Breaking Boundaries in 2024: Stats

    In 2024 the travel sector is flourishing, with nine out of the last 10 record-setting spending days in the global cruise and airline industry occurring this year. Further, travelers are also ...

  18. How much data do you need when traveling internationally?

    As mentioned, how much data you need depends on your general usage patterns and the duration of your travels. To help you estimate your data needs, we went on a five-day trip, and here is a breakdown of our data usage: Total data consumed: 4.07GB. Social Media (Facebook, Instagram, X): 2.3GB. Messaging Apps and Emails: 653MB.

  19. Cross-Country Analysis of Tourist Activities Based on Venue-Referenced

    VR-SMD are similar to georeferenced social media data, such as geotagged travel photos (Vu et al. 2015) and geotagged tweets (Chua et al. 2016), because of their capability to pinpoint the locations of travelers. However, the major advantage of VR-SMD is their capability to associate traveler locations with specific venues rather than the raw ...

  20. Big Data for Better Tourism Policy, Management, and Sustainable

    It highlights how big data is being leveraged for COVID-19 recovery and examines its relationship with statistical frameworks to better measure the economic, social, and environmental impact of tourism. Case studies of partnerships in Asia and the Pacific between the public and private sector demonstrate ways to tap big data.

  21. 12 Ways To Keep Your Personal Data Safe While Traveling

    Odds are good that your email, social media accounts, and many apps have asked you to enable two-factor authentication. Make sure you do so before traveling so you can protect your data.

  22. AI can make planning travel easier, but not without humans

    1:21. It wasn't so long ago that travelers planned trips without the internet. "Back in the day, our parents used to go to these travel agents and really kind of express what they were looking ...

  23. 13 North American amusement parks growing the most in popularity

    To showcase the parks that saw the biggest increases in annual visitorship between 2021 and 2022, Stacker looked at data from the Themed Entertainment Association and AECOM. Read on to find out which 13 North American amusement parks have grown the most in popularity in recent years. 1 / 13. Darren Walsh/Chelsea FC // Getty Images.

  24. CBP Releases April 2024 Monthly Update

    Release Date. Wed, 05/15/2024. WASHINGTON — U.S. Customs and Border Protection (CBP) released operational statistics today for April 2024. CBP monthly reporting can be viewed on CBP's Stats and Summaries webpage. "CBP continues to surge resources and personnel to impacted sectors along the border to ensure the safe, swift, and orderly ...

  25. Chart: Travel becomes more important post-pandemic

    Sep 11, 2023. In a 2022 survey conducted by YouGov, 35% of travellers stated that travel has become more important to them since the pandemic and more than two thirds were planning a trip in the ...

  26. Court Appears Ready to Clear Cruise Lines in Cuba Travel Suit

    Court Appears Ready to Clear Cruise Lines in Cuba Travel Suit. The Eleventh Circuit Friday seemed open to absolving four cruise lines of $440 million in liability for committing "trafficking acts" by carrying passengers to Havana and using port facilities that had been confiscated by Fidel Castro's government in 1960. Plaintiff Havana ...

  27. What is Big Data Analytics?

    The sheer volume of data generated today, from social media feeds, IoT devices, transaction records and more, presents a significant challenge. Traditional data storage and processing solutions are often inadequate to handle this scale efficiently. Big data technologies and cloud-based storage solutions enable organizations to store and manage ...

  28. The giant solar storm is having measurable effects on Earth : NPR

    Around the world on social media, posters put up photos of bright auroras visible in Russia, Scandinavia, the United Kingdom and continental Europe.Some reported seeing the aurora as far south as ...