Exploring the capacity of social media data for modelling travel behaviour: Opportunities and challenges

In the past few years, the social science literature has shown significance attention to extracting information from social media to track and analyse human movements. In this paper the transportation aspect of social media is investigated and reviewed. A detailed discussion is provided about how social media data from different sources can be used to indirectly and with minimal cost extract travel attributes such as trip purpose, mode of transport, activity duration and destination choice, as well as land use variables such as home, job and school location and socio-demographic attributes including gender, age and income. The evolution of the field of transport and travel behaviour around applications of social media over the last few years is studied. Further, this paper presents results of a qualitative survey from travel demand modelling experts around the world on applicability of social media data for modelling daily travel behaviour. The result of the survey reveals positive view of the experts about usefulness of such data sources.     

A coupled multi-agent microsimulation of social interactions and transportation behavior

Choice set formation, location and mode preferences, coordinated scheduling, alternative utility valuations, and shared mobility resources are among the many activity-travel issues hypothesized to be significantly influenced by traveler interdependencies. Empirical evidence lags theory, particularly about the geography of social networks. A simulation tool is presented to let the experimenter construct and test hypothetical interdependencies between geography, socially-linked travelers, and activity-travel choices. The exploratory tool is integrated in the Multi-Agent Transportation Simulation Toolbox (MatSim-T). Initially, any social network can be constructed and embedded in geography. It can remain static, or be adapted to the travel patterns of the agents. The interactions and exchanges between agents influencing socializing and/or travel behavior can be defined in substance and in time/space. The reward for socializing or being socially linked can be varied. Finally, the co-dependence of social factors and travel behavior can be studied. This paper introduces the model and presents verification results which illustrate the coupling of extremely simplified socializing assumptions and travel behavior.     

The impacts of Internet use upon activity participation and travel: Results from a longitudinal diary-based panel study

At the outset of this research, two fields of transport research were in the ascendant. The first reflected technological change, considering the growth of new information and communications technologies and their impacts for transport systems and travel behaviour. The second reflected social change, considering the growth in inequality and disadvantage and the contribution of transport systems and travel behaviour to the same. This paper investigates a potential link between the two, exploring the hypothesis that virtual mobility, via the Internet, could provide a viable alternative to physical mobility in reducing mobility-related social exclusion. The paper presents data from a longitudinal, panel-based diary study. Results support the hypothesis that virtual mobility can provide a viable alternative to physical mobility in reducing aspects of mobility-related exclusion, by providing additional accessibility (virtual accessibility) without an increase in physical mobility. Furthermore, there is no evidence in this research to support a link between physical mobility and virtual mobility; and no evidence to suggest a negative effect of virtual mobility for sociability.     

Identifying tourists and analyzing spatial patterns of their destinations from location-based social media data

Reliable travel behavior data is a prerequisite for transportation planning process. In large tourism dependent cities, tourists are the most dynamic population group whose size and travel choices remain unknown to planners. Traditional travel surveys generally observe resident travel behavior and rarely target tourists. Ubiquitous uses of social media platforms in smartphones have created a tremendous opportunity to gather digital traces of tourists at a large scale. In this paper, we present a framework on how to use location-based data from social media to gather and analyze travel behavior of tourists. We have collected data of about 67,000 users from Twitter using its search interface for Florida. We first propose several filtering steps to create a reliable sample from the collected Twitter data. An ensemble classification technique is proposed to classify tourists and residents from user coordinates. The accuracy of the proposed classifier has been compared against the state-of-the-art classification methods. Finally, different clustering methods have been used to find the spatial patterns of destination choices of tourists. Promising results have been found from the output clusters as they reveal most popular tourist spots as well as some of the emerging tourist attractions in Florida. Performance of the proposed clustering techniques has been assessed using internal clustering validation indices. We have analyzed temporal patterns of tourist and resident activities to validate the classification of the users in two separate groups of tourists and residents. Proposed filtering, identification, and clustering techniques will be significantly useful for building individual-level tourist travel demand models from social media data.     

Statistical and activity-based modeling of university student travel behavior

Abstract

This paper conducts a statistical analysis of student travel behavior at Virginia Commonwealth University (VCU). The data source is the ‘University NHTS’ project launched by the Virginia Department of Transportation (VDOT) in 2009. Through this empirical study, it has been found that university student travel behavior is different from that of the general population; urban universities have lower percentages of nonmotorized trips than college-town universities; undergraduate students are likely to make more daily trips than graduate students – similarly, on-campus students make more frequent trips than off-campus students; the most frequent student activities are home and academic activities; and student group categories have virtually no impact on daily activity profiles, though activity types do have a dramatic impact on daily activity profiles. Based on these research findings, the paper makes a series of recommendations regarding trip generation, trip distribution, mode choice, and activity-based modeling.     

Exploring the role of social capital influence variables on travel behaviour

This paper explores the potential role of individual trip characteristics and social capital network variables in the choice of transport mode. A sample of around 100 individuals living or working in one suburb of Madrid (i.e. Las Rosas district of Madrid) participated in a smartphone short panel survey, entering travel data for an entire working week. A Mixed Logit model was estimated with this data to analyze shifts to metro as a consequence of the opening of two new stations in the area. Apart from classical explanatory variables, such as travel time and cost, gender, license and car ownership, the model incorporated two “social capital network” variables: participation in voluntary activities and receiving help for various tasks (i.e. child care, housekeeping, etc.). Both variables improved the capacity of the model to explain transport mode shifts. Further, our results confirm that the shift towards metro was higher in the case of people “helped” and lower for those participating in some voluntary activities.     

Modeling the behavioral determinants of travel behavior: An application of latent transition analysis

This paper applies the relatively new method of latent class transition analysis to explore the notion that qualitative differences in travel behavior patterns are substantively meaningful and therefore relevant from explanatory point of view. For example, because the bicycle may function as an important access and egress mode, a car user who also (occasionally) uses the bicycle may be more likely to switch to a public transit profile than someone who only uses the car. Data from the Dutch mobility panel are used to inductively reveal travel behavior patterns and model transitions in these patterns over time. Additionally, the effects of seven exogenous variables, including two important life events (i.e. moving house and changing jobs), on cluster membership and the transition probabilities are assessed. The results show that multiple-mode users compared to single-mode users are more likely to switch from one behavioral profile to another. In addition, age, the residential environment, moving house and changing jobs have strong influences on the transition probabilities between the revealed behavioral patterns over time.     

The built environment and travel behavior in urban China: A literature review

Interests in studying of the built environment impacts on travel behavior have proliferated from North America to other parts of the world including China. Until very recently, there has been very little research into travel behavior in China. However, during the last decade, there has been a fast growing interest in studying the built environment and travel behavior in Chinese cities, perhaps motivated by China’s unprecedented urbanization and rapid urban transport development. Case studies from China provide new insights into the impacts of built environment on travel behavior that can help to enrich existing scholarship. However, currently there is a generally poor understanding of the role played by Chinese research and how it has enriched the international literature. This paper aims to fill this gap by reviewing studies in and outside China by both Chinese and non-Chinese scholars. The focus is on the contribution of these studies to the international literature. We identify four areas of contribution: how the built environment has been developed and its implications for travel behavior; the importance of housing sources in defining residential built environment and explaining travel behavior; the unique Danwei (or work unit) perspective on jobs-housing relationships and commuting behavior; and the importance of neighborhood types in explaining travel behavior in Chinese cities. The findings from this review should be relevant for researchers interested in developing future studies that will further advance geographic knowledge of the built environment and travel behavior, specifically in China and with broader global contexts.     

The socio-economics of travel behavior and environmental burdens: A Detroit,Michigan regional context

The transportation industry—particularly light-duty vehicles—is a significant contributor of greenhouse gasses, accounting for about one-third of overall emissions in the U.S. Research to date has studied various factors that impact travel behavior of residents with varying socio-economic characteristics. However, research on the socio-economic characteristics of residents and their impact on environmental burdens within a single urban region, as measured by fuel consumption and vehicular emissions, is recognized as under-represented in the U.S. planning and transportation literature. This study focuses on the Detroit region, Michigan, a unique case study due to the scale of suburbanization and urban decline, yet representative of many mid-western cities. The article explores how socio-economic characteristics impact travel patterns and environmental burdens within six Detroit region neighborhoods. Data on individual travel behavior and personal vehicle characteristics gathered from a mail survey enabled an analysis into how associated environmental burdens varied with socio-economic composition. The analysis explores contributions to environmental burdens between poorer urban and wealthier suburban populations.     

From Twitter to detector: Real-time traffic incident detection using social media data

The effectiveness of traditional incident detection is often limited by sparse sensor coverage, and reporting incidents to emergency response systems is labor-intensive. We propose to mine tweet texts to extract incident information on both highways and arterials as an efficient and cost-effective alternative to existing data sources. This paper presents a methodology to crawl, process and filter tweets that are accessible by the public for free. Tweets are acquired from Twitter using the REST API in real time. The process of adaptive data acquisition establishes a dictionary of important keywords and their combinations that can imply traffic incidents (TI). A tweet is then mapped into a high dimensional binary vector in a feature space formed by the dictionary, and classified into either TI related or not. All the TI tweets are then geocoded to determine their locations, and further classified into one of the five incident categories.We apply the methodology in two regions, the Pittsburgh and Philadelphia Metropolitan Areas. Overall, mining tweets holds great potentials to complement existing traffic incident data in a very cheap way. A small sample of tweets acquired from the Twitter API cover most of the incidents reported in the existing data set, and additional incidents can be identified through analyzing tweets text. Twitter also provides ample additional information with a reasonable coverage on arterials. A tweet that is related to TI and geocodable accounts for approximately 5% of all the acquired tweets. Of those geocodable TI tweets, 60–70% are posted by influential users (IU), namely public Twitter accounts mostly owned by public agencies and media, while the rest is contributed by individual users. There is more incident information provided by Twitter on weekends than on weekdays. Within the same day, both individuals and IUs tend to report incidents more frequently during the day time than at night, especially during traffic peak hours. Individual tweets are more likely to report incidents near the center of a city, and the volume of information significantly decays outwards from the center.     

Exploring the propensity to perform social activities: a social network approach

Conceptual and empirical models of the propensity to perform social activity–travel behavior are described, which incorporate the influence of individuals’ social context, namely their social networks. More explicitly, the conceptual model develops the concepts of egocentric social networks, social activities, and social episodes, and defines the three sets of aspects that influence the propensity to perform social activities: individuals’ personal attributes, social network composition, and information and communication technology interaction with social network members. Using the structural equation modeling (SEM) technique and data recently collected in Toronto, the empirical model tests the effect of these three aspects on the propensity to perform social activities. Results suggest that the social networks framework provides useful insights into the role of physical space, social activity types, communication and information technology use, and the importance of “with whom” the activity was performed with. Overall, explicitly incorporating social networks into the activity–travel behavior modeling framework provides a promising framework to understand social activities and key aspects of the underlying behavioral process. Juan Antonio Carrasco a PhD candidate in Civil Engineering at the University of Toronto, holds a MSc degree in Transportation Engineering from the Pontificia Universidad Católica de Chile. His doctoral research explores the relationships between social networks, activity–travel behavior, and ICTs. His research interests also include microsimulation, land use-transportation, and econometric modeling. Eric J. Miller is Bahen-Tanenbaum Professor of Civil Engineering at the University of Toronto where he is also Director of the Joint Program in Transportation. His research interests include integrated land-use/transportation modeling, activity-based travel modeling, microsimulation and sustainable transportation planning.     

Illuminating the unseen in transit use: A framework for examining the effect of attitudes and perceptions on travel behavior

This study develops the Perception–Intention–Adaptation (PIA) framework to examine the role of attitudes, perceptions, and norms in public transportation ridership. The PIA framework is then applied to understand the relative importance of socio-demographic, built environment, transit service, and socio-psychological factors on public transit use for 279 residents of south Los Angeles, California, a predominately low-income, non-white neighborhood. Confirmatory factor analysis based on 21 survey items resulted in six transit-relevant socio-psychological factors which were used in regression models of two measures of transit use: the probability of using transit at least once in the 7-day observation period, and the mean number of daily transit trips. Our analysis indicates that two PIA constructs, attitudes toward public transportation and concerns about personal safety, significantly improved the model fit and were robust predictors of transit use, independent of built environment factors such as near-residence street network connectivity and transit service level. Results indicate the need for combined policy approaches to increasing transit use that not only enhance transit access, but also target attitudes about transit service and perceptions of crime on transit.     

The cost of equity: Assessing transit accessibility and social disparity using total travel cost

Social equity is increasingly incorporated as a long-term objective into urban transportation plans. Researchers use accessibility measures to assess equity issues, such as determining the amount of jobs reachable by marginalized groups within a defined travel time threshold and compare these measures across socioeconomic categories. However, allocating public transit resources in an equitable manner is not only related to travel time, but also related to the out-of-pocket cost of transit, which can represent a major barrier to accessibility for many disadvantaged groups. Therefore, this research proposes a set of new accessibility measures that incorporates both travel time and transit fares. It then applies those measures to determine whether people residing in socially disadvantaged neighborhoods in Montreal, Canada experience the same levels of transit accessibility as those living in other neighborhoods. Results are presented in terms of regional accessibility and trends by social indicator decile. Travel time accessibility measures estimate a higher number of jobs that can be reached compared to combined travel time and cost measures. However, the degree and impact of these measures varies across the social deciles. Compared to other groups in the region, residents of socially disadvantaged areas have more equitable accessibility to jobs using transit; this is reflected in smaller decreases in accessibility when fare costs are included. Generating new measures of accessibility combining travel time and transit fares provides more accurate measures that can be easily communicated by transportation planners and engineers to policy makers and the public since it translates accessibility measures to a dollar value.     

Assessing social equity in distance based transit fares using a model of travel behavior

The goal of this study is to develop and apply a new method for assessing social equity impacts of distance-based public transit fares. Shifting to a distance-based fare structure can disproportionately favor or penalize different subgroups of a population based on variations in settlement patterns, travel needs, and most importantly, transit use. According to federal law, such disparities must be evaluated by the transit agency, but the area-based techniques identified by the Federal Transit Authority for assessing discrimination fail to account for disparities in distances travelled by transit users. This means that transit agencies currently lack guidelines for assessing the social equity impacts of replacing flat fare with distance-based fare structures. Our solution is to incorporate a joint ordinal/continuous model of trip generation and distance travelled into a GIS Decision Support System. The system enables a transit planner to visualize and compare distance travelled and transit-cost maps for different population profiles and fare structures. We apply the method to a case study in the Wasatch Front, Utah, where the Utah Transit Authority is exploring a switch to a distance-based fare structure. The analysis reveals that overall distance-based fares benefit low-income, elderly, and non-white populations. However, the effect is geographically uneven, and may be negative for members of these groups living on the urban fringe.     

Investigating travel utility elements in association with travel time and mode choice: the case of Seoul,South Korea

In this paper, travel utility is conceptualized into the elements of disutility, or derived utility, and positive utility, which includes synergistic and intrinsic utility, and then analyzed in terms of the effects of these elements on weekly travel time according to three travel modes – the automobile, public transit, and nonmotorized modes – and on the choice of the annually most used mode. Linear regressions on mode-specific travel time and a multinomial logistic regression on mode choice show that, compared to life situation and land-use characteristics, utility elements are among the strongest travel determinants. Specifically, while some utility elements contribute exclusively to shifting the mode of travel and others to increasing nonmotorized travel, modal shift is most strongly affected by a disutility element, trip timeliness, and the increase in nonmotorized travel by a positive utility element, amenities.     

Rewarding rush-hour avoidance: A study of commuters’ travel behavior

Spitsmijden, peak avoidance in Dutch, is the largest systematic effort to date to study, in the field, the potential of rewards as a policy mean for changing commuter behavior. A 13 week field study was organized in The Netherlands with the purpose of longitudinally investigating the impacts of rewards on commuter behavior. Different levels and types of rewards were applied and behavior was tracked with state-of-the art detection equipment. Based on the collected data, which included also pre and post-test measurements, a mixed discrete choice model was estimated. The results suggest that rewards can be effective tools in changing commuting behavior. Specifically rewards reduce the shares of rush-hour driving, shift driving to off-peak times and increase the shares of public transport, cycling and working from home. Mediating factors include socio-demographic characteristics, scheduling constraints and work time flexibility, habitual behavior, attitudes to commuting alternatives, the availability of travel information and even the weather. The success of this study has encouraged adoption of rewards, as additional policy tools, to alleviate congestion, especially during temporary road closures.     

Reintroducing attitude theory in travel behavior research: The validity of an interactive interview procedure to predict car use

A methodological challenge is to develop methods which satisfy the need in transport planning of accurately forecasting travel behavior. Drawing on a review of the current state of attitude theory, it is argued that successfully forecasting travel behavior relies on a distinction between planned, habitual, and impulsive travel. Empirical illustrations are provided in the form of stated-response data from two experiments investigating the validity of an interactive interview procedure to predict household car use for different types of trips, either before or after participants were required to reduce use.     

Toward using social media to support ridesharing services: challenges and opportunities

Ridesharing has been attracting increasing attention from both academia and industry. One of the challenges posed by the study of ridesharing is to identify the most valuable information for improving the ridesharing decisions taken by participants. Another challenge is to use harvesting techniques to extract specific types of travel-related information. Many methods have been developed by the community in order to solve these issues. However, due to a lack of information sharing between different transit authorities and the difficulty of identifying subjective perceptions of the experience of ridesharing, understanding and evaluating how social media data might support or obstruct goals for mobility, safety and environmental sustainability in ridesharing is a difficult task. In this survey, we first analyze the literature on ridesharing with a focus on the properties and model of service, and introduce a framework to describe the major components required for a ridesharing service. Then, we illustrate the potential value of information extracted from social media and present the rationale for harvesting travel-related data. Finally, we detail some possible directions and different approaches for using social media data, and highlight their assets and drawbacks.