Geotagged Twitter Data Research Articles

Geolocated Twitter-based population mobility in Victoria, Australia, during the staged COVID-19 restrictions

Using geotagged Twitter data in Victoria, we created a mobility index and studied the changes during the staged restrictions during the coronavirus disease 2019 (COVID-19) pandemic. We describe preliminary evidence that geotagged Twitter data may be used to provide real-time population mobility data and information on the impact of restrictions on such mobility.

Using Social Media to Predict Food Deserts in the United States: Infodemiology Study of Tweets

BackgroundThe issue of food insecurity is becoming increasingly important to public health practitioners because of the adverse health outcomes and underlying racial disparities associated with insufficient access to healthy foods. Prior research has used data sources such as surveys, geographic information systems, and food store assessments to identify regions classified as food deserts but perhaps the individuals in these regions unknowingly provide their own accounts of food consumption and food insecurity through social media. Social media data have proved useful in answering questions related to public health; therefore, these data are a rich source for identifying food deserts in the United States.ObjectiveThe aim of this study was to develop, from geotagged Twitter data, a predictive model for the identification of food deserts in the United States using the linguistic constructs found in food-related tweets.MethodsTwitter’s streaming application programming interface was used to collect a random 1% sample of public geolocated tweets across 25 major cities from March 2020 to December 2020. A total of 60,174 geolocated food-related tweets were collected across the 25 cities. Each geolocated tweet was mapped to its respective census tract using point-to-polygon mapping, which allowed us to develop census tract–level features derived from the linguistic constructs found in food-related tweets, such as tweet sentiment and average nutritional value of foods mentioned in the tweets. These features were then used to examine the associations between food desert status and the food ingestion language and sentiment of tweets in a census tract and to determine whether food-related tweets can be used to infer census tract–level food desert status.ResultsWe found associations between a census tract being classified as a food desert and an increase in the number of tweets in a census tract that mentioned unhealthy foods (P=.03), including foods high in cholesterol (P=.02) or low in key nutrients such as potassium (P=.01). We also found an association between a census tract being classified as a food desert and an increase in the proportion of tweets that mentioned healthy foods (P=.03) and fast-food restaurants (P=.01) with positive sentiment. In addition, we found that including food ingestion language derived from tweets in classification models that predict food desert status improves model performance compared with baseline models that only include socioeconomic characteristics.ConclusionsSocial media data have been increasingly used to answer questions related to health and well-being. Using Twitter data, we found that food-related tweets can be used to develop models for predicting census tract food desert status with high accuracy and improve over baseline models. Food ingestion language found in tweets, such as census tract–level measures of food sentiment and healthiness, are associated with census tract–level food desert status.

Insights from the relationship between urban form, social media, and edu-tourism

ABSTRACT Educational tourism and urban spaces are complex and interwoven phenomena of the contemporary globalized world. The extended windows of time that students of higher education institutes spend in the host cities makes them more than mere visitors; they become part of the everyday life of the urban context. Nevertheless, the interdisciplinary relationship between urban studies and edu-tourism remains understudied in contemporary literature, especially considering emerging types of data that can provide new insights. This paper draws on volunteered geographic information to explore interactions between higher education students and their host cities. Geotagged Twitter data was analysed in terms of both spatial density and content. The study was conducted in two coastal cities of Cyprus. The analysis indicates tendencies for student interactions with spaces outside the university campuses—with the majority of tweets associated with non-educational interaction types and venues with diverse spatial signatures. This study argues that edu-tourism is largely associated with urban tourism and it is essential to consider these interactions in decision-making and urban planning to improve both the tourism industry and the urban spaces.

Human mobility forecasting with region-based flows and geotagged Twitter data

One of the main lines of research in the discipline of mobility mining is the development of predictors able to anticipate human travel behaviour in great detail. However, access to the high-resolution spatio-temporal data on which most existing solutions are based is rather limited due to multiple factors, e.g. costly access to third-party data. These restrictions give rise to a problem of developing predictors of human mobility in most setting, since the amount of data available to train these prediction models is insufficient. This paper explores the feasibility of using a public data source such as Twitter to predict the number of trips at the nationwide level. The proposed approach combines a large set of geotagged Twitter posts with an open data source published by the Spanish government on traveller mobility based on mobile phone location. Both datasets are used as input to Machine Learning models to validate the use of Twitter data for improving the prediction of these models. The results show that Twitter data have considerable value as a predictor of large-scale human mobility, especially for Long Short-Term Memory (LSTM) models. As a result, the relevance of this work resides in demonstrating that the use of Twitter could be considered as an alternative to substantially enhance the prediction of mobility within a country when it is combined with other open data sources.

A Taxi Trajectory and Social Media Data Management Platform for Tourist Behavior Analysis

Taxis are commonly used by tourists to travel around unfamiliar cities they visit. These taxis today have GPS devices, which can then be used to collect a significant amount of data on the movement of tourists. One problem with this idea, however, is the question of how to extract that movement data from the raw GPS data, which includes a lot of other data, such as vehicle IDs, timestamps, and speeds, etc. The purpose of this research is to propose a data management platform to process heterogeneous data including taxi data, social media data, and place data for tourist behavior analysis. We propose a data pipeline that can be scaled in order to process a significant amount of data regarding taxi trajectory and social media, with two objectives. The first objective is to extract the tourist trajectory data from the raw GPS data and produce a data integration module enriched with a knowledge base of tourist trajectories. This knowledge base is constructed through the extension of semantic trajectory ontology (STO) and mobility behavior ontology (MBO). The second objective is to extract tourist activities/point of interests (POIs) from geo-tagged Twitter data. The results of the data pipeline can readily be used for tourist behavior analysis, such as tourist descriptive analysis, popular tourist destinations/zones, and tourist movement patterns identification. We leverage the study’s results to demonstrate the real-life case study in Bangkok during the Songkran Festival in 2019. Thus, we could precisely identify tourist movement during various periods, determine popular destinations/zones, discover high density density of taxi destination points for a given trajectory type, and display the top ten tourist destinations, as well as prominent tourism keywords or trends at the time. This can provide insight to governments and businesses related to tourism regarding the trajectories and activities of tourists, and it will help predict future tourism trends.

The promise of excess mobility analysis: measuring episodic-mobility with geotagged social media data

ABSTRACT Human mobility studies have become increasingly important and diverse in the past decade with the support of social media big data that enables human mobility to be measured in a harmonized and rapid manner. However, what is less explored in the current scholarship is episodic mobility as a special type of human mobility defined as the abnormal mobility triggered by episodic events excess to the normal range of mobility at large. Drawing on a large-scale systematic collection of 1.9 billion geotagged Twitter data from 2017 to 2020, this study contributes to the first empirical study of episodic mobility by producing a daily Twitter census of visitors at the U.S. county level and proposing multiple statistical approaches to identify and quantify episodic mobility. It is followed by four case studies of episodic mobility in U.S. national wide to showcase the great potential of Twitter data and our proposed method to detect episodic mobility subject to episodic events that occur both regularly and sporadically. This study provides new insights on episodic mobility in terms of its conceptual and methodological framework and empirical knowledge, which enriches the current mobility research paradigm.

Analysis of Geotagging Behavior: Do Geotagged Users Represent the Twitter Population?

Twitter’s APIs are now the main data source for social media researchers. A large number of studies have utilized Twitter data for diverse research interests. Twitter users can share their precise real-time location, and Twitter APIs can provide this information as longitude and latitude. These geotagged Twitter data can help to study human activities and movements for different applications. Compared to the mostly small-scale data samples in different domains, such as social science, collecting geotagged data offers large samples. There is a fundamental question whether geotagged users can represent non-geotagged users. While some studies have investigated the question from different perspectives, they did not investigate profile information and the contents of tweets of geotagged and non-geotagged users. This empirical study addresses this limitation by applying text mining, statistical analysis, and machine learning techniques on Twitter data comprising more than 88,000 users and over 170 million tweets. Our findings show that there is a significant difference (p-value < 0.001) between geotagged and non-geotagged users based on 73% of the features obtained from the users’ profiles and tweets. The features can also help to distinguish between geotagged and non-geotagged users with around 80% accuracy. This research illustrates that geotagged users do not represent the Twitter population.

Analysis of mobility homophily in Stockholm based on social network data.

We present a novel metric for measuring relative connection between parts of a city using geotagged Twitter data as a proxy for co-occurrence of city residents. We find that socioeconomic similarity is a significant predictor of this connectivity metric, which we call “linkage strength”: neighborhoods that are similar to one another in terms of residents’ median income, education level, and (to a lesser extent) immigration history are more strongly connected in terms of the of people who spend time there, indicating some level of homophily in the way that individuals choose to move throughout a city’s districts.

Social distancing beliefs and human mobility: Evidence from Twitter.

We construct a novel database containing hundreds of thousands geotagged messages related to the COVID-19 pandemic sent on Twitter. We create a daily index of social distancing—at the state level—to capture social distancing beliefs by analyzing the number of tweets containing keywords such as “stay home”, “stay safe”, “wear mask”, “wash hands” and “social distancing”. We find that an increase in the Twitter index of social distancing on day t-1 is associated with a decrease in mobility on day t. We also find that state orders, an increase in the number of COVID-19 cases, precipitation and temperature contribute to reducing human mobility. Republican states are also less likely to enforce social distancing. Beliefs shared on social networks could both reveal the behavior of individuals and influence the behavior of others. Our findings suggest that policy makers can use geotagged Twitter data—in conjunction with mobility data—to better understand individual voluntary social distancing actions.

Changing neighborhoods, shifting connections: mapping relational geographies of gentrification using social media data

ABSTRACT New sources of ‘big data’ are regularly described as revolutionizing the study of urban life. Of particular interest is analyzing gentrification, which has proven a challenging endeavor with conventional methods. Big data may offer a new approach to the persistent problem of defining and measuring gentrification, while also allowing us to rethink broader questions about theory and methodology in urban geography. Using geotagged Twitter data, we demonstrate how the changing geographies of users’ tweets are proxies for the evolving social and spatial contours of urban neighborhoods. We use the case of Lexington, Kentucky to analyze the mobilities and relational connections of neighborhood residents and visitors as gentrification intensified over time. We argue that these kinds of big data allow for an analytical approach that focuses on the dynamic, relational connections between people and places, and provides a useful, additional avenue in understanding a process as complex and multifaceted as gentrification.

How does socio-economic and demographic dissimilarity determine physical and virtual segregation?

It is established that socio-economic and demographic dissimilarities between populations are determinants of spatial segregation. However, the understanding of how such dissimilarities translate into actual segregation is limited. We propose a novel network-analysis approach to comprehensively study the determinants of communicative and mobility-related spatial segregation, using geo-tagged Twitter data. We constructed weighted spatial networks representing tie strength between geographical areas, then modeled tie formation as a function of socio-economic and demographic dissimilarity between areas. Physical and virtual tie formation were affected by income, age, and race differences, although these effects were smaller by an order of magnitude than the geographical distance effect. Tie formation was more frequent when "destination" area had higher median income and lower median age. We hypothesize that physical tie formation is more "costly" than a virtual one, resulting in stronger segregation in the physical world. Economic and cultural motives may result in stronger segregation of relatively rich and young populations from their surroundings. Our methodology can help identify types of states that lead to spatial segregation, and thus guide planning decisions for reducing its adverse effects.

TwitterデータからみたCOVID-19の拡大に対する「行動変容」

TwitterデータからみたCOVID-19の拡大に対する「行動変容」

A general approach to detecting migration events in digital trace data.

Empirical research on migration has historically been fraught with measurement challenges. Recently, the increasing ubiquity of digital trace data—from mobile phones, social media, and related sources of ‘big data’—has created new opportunities for the quantitative analysis of migration. However, most existing work relies on relatively ad hoc methods for inferring migration. Here, we develop and validate a novel and general approach to detecting migration events in trace data. We benchmark this method using two different trace datasets: four years of mobile phone metadata from a single country’s monopoly operator, and three years of geo-tagged Twitter data. The novel measures more accurately reflect human understanding and evaluation of migration events, and further provide more granular insight into migration spells and types than what are captured in standard survey instruments.

Reading the urban socio-spatial network through space syntax and geo-tagged Twitter data

ABSTRACT This research examines the predictability of geo-tagged Twitter data via space syntax integration measurement using geographic information system (GIS). Mobile networking mediums are an emergent part of everyday life in the city. In light of this, research on cities needs to take into account these new data sources. The study was conducted in two Cypriot cities, Famagusta and Kyrenia, and the result shows the significance of local accessibility in predicting Twitter data in both cases. The outcome also suggests the critical importance of investigating the outliers in the dataset because they might clarify hidden potentials of urban spaces.

Geolocated Twitter social media data to describe the geographic spread of SARS-CoV-2.

Openly available, geotagged Twitter data from 2013 to 2015 was used to estimate the 2019–2020 human mobility patterns in and outside of China to predict the spatiotemporal spread of severe acute respiratory syndrome coronavirus 2. Countries with the highest number of visiting Twitter users outside of China were the USA, Japan, UK, Germany and Turkey. A high correlation was observed when comparing country-level Twitter user visits and reported cases.

Characterizing traveling fans: a workflow for event-oriented travel pattern analysis using Twitter data

ABSTRACT Characterizing event attendees’ travel patterns is key to understanding the dynamics of social events in cities. However, the scientific investigation of event travel patterns has been hindered by the difficulty in gathering travel diaries of participants. Geotagged microblogs provide new opportunities for studying event travel patterns by offering rich locational and semantic information of attendees. Here, we develop, implement, and apply a workflow to characterize travel behaviors of event attendees with geotagged Twitter data, using college football events as a case study. The workflow includes five steps: 1) filtering event attendees using real-time geotagged tweets, 2) identifying origins of the event attendees using historical timeline tweets, 3) identifying past sports-related activities at travel destinations using topic modeling, 4) computing user movement features using origin-destination travel flows, and 5) identifying atypical travel patterns to characterize event attendees. The travel patterns uncovered in the study offer insights into user interests and travel behaviors related to sporting event attendance. The findings demonstrate that our method holds promise in revealing long-term event travel patterns (not limited to sporting events) through the use of geotagged microblogs.

Using socially-sensed data to infer ZIP level characteristics for the spatiotemporal analysis of drug-related health problems in Maryland.

This research investigated how socially sensed data can be used to detect ZIP level characteristics that are associated with spatial and temporal patterns of Emergency Department patients with a chief complaint and/or diagnosis of overdose or drug-related health problems for four hospitals in Baltimore and Anne Arundel County, MD during 2016-2018. Dynamic characteristics were identified using socially-sensed data (i.e., geo-tagged Twitter data) at ZIP code level over varying temporal resolutions. Data about three place-based variables including comments and concerns about crime, drug use, and negative or depressed sentiments, were extracted from tweets, along with data from four socio-environmental variables from the American Community Survey were collected to explore socio-environmental characteristics during the same period. Our study showed a statistically significant increase in adjusted rates of Emergency Department (ED) visits occurred between June and November 2017 for patients residing in ZIP codes in western Baltimore and northeastern Anne Arundel County. During this period, the three topics extracted from Twitter data were highly correlated with the ZIP codes where the patients were residing. Exploring the dynamic spatial associations between socio-environmental variables and ED visits for acute overdose assists local health officials in optimizing interventions for vulnerable locations.

Examine the effects of neighborhood equity on disaster situational awareness: Harness machine learning and geotagged Twitter data

A disaster-resilient city should address social inequity in all its forms. Complete, accurate, and up-to-the-minute situational awareness (SA) can help disaster relief organizations stabilize the dangers and prevent further losses in poor and ethnic minority neighborhoods. However, the lapse of months or years of survey data could lead to biases since retrospective studies are affected by the emotional status at the time of the survey. To address this, we aim to examine the effects of neighborhood equity on SA in a hurricane event using geotagged Twitter data. We adopted the sentiment analysis, convolutional neural network (CNN) model and latent Dirichlet allocation (LDA) model to reflect SA in Hurricane Florence. The multinomial logit regression model reveals that the tweet originated from a poor neighborhood has a higher probability of being more negative, compared to being neutral. Also, we surprisingly found that the sentiment of a tweet in a black neighborhood could be less likely to be negative. Another novel finding is that black neighborhoods could discuss the hurricane with a positive attitude while poor neighborhoods are more concerned about the work during the hurricane. This study shows that, by incorporating with aggregated sociodemographic data, geotagged Twitter data can also be used to understand disaster SA from the perspective of social equity.

Delineating and modeling activity space using geotagged social media data

ABSTRACTIt has become increasingly important in spatial equity studies to understand activity spaces – where people conduct regular out-of-home activities. Big data can advance the identification of activity spaces and the understanding of spatial equity. Using the Los Angeles metropolitan area for the case study, this paper employs geotagged Twitter data to delineate activity spaces with two spatial measures: first, the average distance between users’ home location and activity locations; and second, the area covered between home and activity locations. The paper also finds significant relationship between the spatial measures of activity spaces and neighborhood spatial and socioeconomic characteristics. This research enriches the literature that aims to address spatial equity in activity spaces and demonstrates the applicability of big data in urban socio-spatial research.

Urban Activity Detection Using Geo-located Twitter Data

More and more studies are based on freely available social media data. Using microblogs, a midpoint between instant messaging and content production, analyses of urban activities are possible. This paper focuses not only on mapping human activities but also on defining urban function in the city. Using geotagged Twitter data, the research carried out separate spatial and temporal analyses, in conjunction with combined spatio-temporal analyses. Tweets were categorised into six activity groups: Working, Eating, Shopping, Leisure, Home and Education, based on selected keywords. The results show stronger performance for the detection of Leisure, Eating, Shopping and Education activities and less successful performance for Working and Home activities. The first four cluster near the centre of the city, while the rest are scattered all over the city. Moreover, each activity shows its own temporal pattern. This study finds characteristic patterns for everyday activities and shows the possibility of using social media data to define urban function for places where land-use information is not available.