Travel Behaviour and Society最新文献

{"title":"Learning nonlinearity and measuring uncertainty--a multi-task neural network and additive gaussian process based travel choice model","authors":"Sha Zhang ,&nbsp;Yao Dong ,&nbsp;Peter J. Jin ,&nbsp;Shichao Sun ,&nbsp;Fei Yang","doi":"10.1016/j.tbs.2026.101247","DOIUrl":"10.1016/j.tbs.2026.101247","url":null,"abstract":"<div><div>Discrete choice model (DCM) is a classical framework for modelling an individual’s travel choice. However, its oversimplified architecture of utility function may limit its performance when faced with a complex decision process. In this paper, we develop a new framework called multi-task neural network and additive gaussian process based discrete choice model (MNNAGP-DCM). Specially, the multi-task neural network (MNN) is used to learn the representation of individual characteristics, while the additive Gaussian process regression (AGP) process is utilized to enhance flexibility of utility function. In multi-task neural network, the sub-learners learn the taste parameters between individuals’ characteristic in each alternative, while the global bias term is used to learn the cross effect between alternatives. The additive GPR framework is employed to substitute the linear term in utility function with a nonparametric probability framework. Additive GPR enables the modelling of nonlinearity, threshold effects and uncertainty, thereby providing a more comprehensive perspective on the decision-making process. Moreover, when combined with DCM, the GPRs become intractable. To address this, we employ variational inference to construct a tractable lower bound, thereby transforming the original model into a tractable one. Then MNNAGP-DCM can be optimized by gradient based algorithms such as Adam. The proposed model is tested on the open-source dataset and benchmarked with standard MNL, Mix-logit, XGBoost, TasteNet-MNL, MNN-DCM and MNNSGP-DCM. Results show that MNNAGP-DCM can not only capture individuals’ heterogeneity but also can learn the nonlinearity in utility function, showing great superiority in terms of predictability. Our model can also provide interpretable result with taste parameters and the fitted GPR models, while quantifying uncertainty through GPR’s probability framework.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"44 ","pages":"Article 101247"},"PeriodicalIF":5.7,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146032756","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Navigating the gig economy: transportation labor challenges facing California’s app-based ridehailing and courier drivers","authors":"Susan Shaheen ,&nbsp;Brooke Wolfe ,&nbsp;Adam Cohen","doi":"10.1016/j.tbs.2025.101218","DOIUrl":"10.1016/j.tbs.2025.101218","url":null,"abstract":"<div><div>Given the dynamic landscape surrounding the classification of workers in California, it is important to consider how the existing legal and regulatory environment may impact app-based gig drivers, including transportation network companies (TNCs, also known as ridehailing) and courier network services (CNS). Using a multi-method approach, we conducted a literature review (n = 41 sources), expert interviews (n = 8), and case study analysis (n = 7) between October 2022 to May 2024 to better understand how California’s gig drivers are impacted by state legislation and regulation (i.e., Assembly Bill 5, Proposition 22, and Senate Bill 1014). The expert interviews found that gig drivers are concerned with fair pay, benefits, labor classification, and transparency from the app-based platforms on issues related to punitive actions (i.e., deactivations). Drivers also raised concerns about California’s Clean Miles Standard (also known as SB 1014), which requires 90 % of vehicle miles traveled be electric by 2030, due to the financial costs associated with acquiring and operating electric vehicles (EVs) and limited public charging availability. The seven case studies examine gig labor policies from other states and countries (e.g., policies that may help enhance driver working conditions, platform regulation, and facilitate the EV transition). Together, these methods explore the tension between sustaining an app-based gig driving platform and providing fair compensation and working conditions for gig drivers. The study finds that state and/or local policies establishing minimum pay for drivers and policies enhancing transparency and appeal processes for driver deactivations could help improve working conditions for gig drivers. Various state agencies, such as the California Public Utilities Commission, could support gig drivers through incentives for the purchase of EVs and installation of EV charging near their homes and driving locations.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"44 ","pages":"Article 101218"},"PeriodicalIF":5.7,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146015817","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Extracting socio-psychological perceptions for analysis of travel behaviours","authors":"Yanyan Xu,&nbsp;Panchamy Krishnakumari,&nbsp;Neil Yorke-Smith,&nbsp;Serge Hoogendoorn","doi":"10.1016/j.tbs.2025.101197","DOIUrl":"10.1016/j.tbs.2025.101197","url":null,"abstract":"<div><div>This article proposes an evidence-based policy recommendation framework integrating social media data and natural language processing methods, to support inclusive and efficient transport policy-making. Given that current research underscores the crucial role of both external and psychological variables in individual travel decisions, psychological features – such as beliefs, attitudes or values – are frequently used as latent variables for travel behaviour interpretation and travel choice modelling. However, user-centric policy recommendations based on dynamic psychological variables are still limited. Most studies rely on survey data, which neglects the urgent dynamic trend of user perception change and its underlying relationship with travel behaviour. Hence there is a lack of illustration on how these psychological variables can be further used at specific temporal and spatial levels for travel behaviour interpretation. This would be valuable to identify priorities for more targeted (sustainability and other) policies and interventions. In this article, we utilize sentiment analysis and dynamic topic modelling to represent the spatial–temporal variance of psychological features. Integrating with corresponding travel behaviour, we illustrate how these dynamic psychological features can distinguish travel dissonance, identify key motivations, and reflect urgent social demands at precise spatial–temporal levels. We demonstrate these advances in a case study in New York City from 2019 to 2022 using Twitter (X) data. A comparison with existing travel-related policies in the case study validates the feasibility of our framework to support evidence-based policy recommendations. We conclude by discussing the potential of this framework to support sustainable transport promotion.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"43 ","pages":"Article 101197"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145732698","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring different patterns of bike-and-ride trips and influencing factors using geographically weighted random forest","authors":"Zijian Yang ,&nbsp;Guocong Zhai ,&nbsp;N.N. Sze ,&nbsp;Hongliang Ding ,&nbsp;Nikolai Bobylev ,&nbsp;Hongtai Yang","doi":"10.1016/j.tbs.2025.101201","DOIUrl":"10.1016/j.tbs.2025.101201","url":null,"abstract":"<div><div>Numerous studies have examined the effectiveness of urban transportation planning policies such as park-and-ride, transit-oriented development, and multimodal transportation hubs in promoting public transit use. In the past decades, bike sharing, both dockless and docking systems, has been increasingly popular as a green transport mode, connecting to public transit. However, the influence of socioeconomic conditions, land-use factors, transport infrastructure, and station characteristics on bike-and-ride remains underexplored, particularly across different patterns of bike-and-ride. In this study, an integrated random forest (RF) and geographically weighted regression (GWR) model is applied to capture nonlinear relationships and spatial heterogeneity between bike-and-ride usage and explanatory variables, including socioeconomic variables, land-use factors, transportation infrastructure, and metro characteristics, based on the integrated bike sharing and metro ridership data in Chengdu, China. Additionally, a novel similarity metric, Dynamic Time Warping (DTW), is applied to classify the metro stations based on the temporal pattern of bike-and-ride trips at all stations. Hence, four clusters of metro stations with a varying pattern of bike-and-ride trips are established. The results show that the importance of determinants and their association with bike-and-ride vary significantly among different clusters of metro stations. This study proposes a fine-grained analytical framework for bike-and-ride, providing theoretical and empirical support for station function classification.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"43 ","pages":"Article 101201"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145731902","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The impact of perceived walking accessibility on willingness to walk: evaluating different assessment methods","authors":"Cong Qi ,&nbsp;Jonas De Vos ,&nbsp;Yibang Zhang ,&nbsp;Xiucheng Guo","doi":"10.1016/j.tbs.2025.101173","DOIUrl":"10.1016/j.tbs.2025.101173","url":null,"abstract":"<div><div>Accurately assessing perceived walking accessibility is essential for analysing its impact on people’s willingness to walk and walking behaviour. However, limited research has analysed and compared different assessments of perceived walking accessibility. This paper uses survey data from Nanjing’s central area to build binary logit models and analyse the impact of perceived walking accessibility on willingness to walk. Four different assessment methods are employed: the perceived accessibility scale, perceived walkability, overall perceived walking accessibility, and perceived impedance. The results show that perceived walking time is the most effective method for assessing perceived walking accessibility to specific locations, with the highest degree of explanatory power for willingness to walk. Older people, those unfamiliar with the city centre and those who arrive there by car or bicycle are less likely to walk in the central area. Perceived walking time is the primary factor influencing both enthusiastic walkers and reluctant walkers, while actual walking time primarily influences conditional walkers. Land use type at survey point has no significant effect on willingness to walk. These findings are valuable for designing an appropriate walking environment in the city centre and for encouraging walking by improving people’s perceived walking accessibility.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"43 ","pages":"Article 101173"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145442553","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Miles with smiles: the role of e-cargo bikes in facilitating new personal and family-oriented travel and relevant beyond-utility motivations","authors":"Labib Azzouz ,&nbsp;Christian Brand ,&nbsp;Noel Cass ,&nbsp;Ian Philips","doi":"10.1016/j.tbs.2025.101217","DOIUrl":"10.1016/j.tbs.2025.101217","url":null,"abstract":"<div><div>E-cargo bikes (ECBs) can play a crucial role in the transition to sustainable transport. Existing research primarily focuses on ECBs in sharing schemes and urban delivery, with limited attention to domestic use. Most studies emphasize mode substitution, often overlooking motivations unique to ECBs and beyond-utility travel motivations. Critically, little is known about ECBs’ role in generating new travel demand. This study explores how ECBs generate new trips, focusing on individual and household motivations that extend beyond purely utilitarian purposes. Trials were conducted with 49 households across three cities: Leeds, Oxford, and Brighton. A mixed-methods approach was employed, emphasizing qualitative data from interviews and supplemented with quantitative insights from travel diaries.</div><div>Findings indicate that ECBs enhanced accessibility, leading to increased travel distance and frequency, and enabling travelers to ‘do more.’ Their capacity to transport children and bulky items unlocked induced and latent demand, facilitating trips that otherwise would not have occurred. Beyond utility, ECBs fostered new solo and family travel shaped by a range of intrinsic motivations. They promoted well-being, offered therapeutic outdoor experiences, disrupted daily routines, and supported personal growth, freedom, and autonomy. Caregivers particularly valued ECBs for the control, spontaneity, and flexibility they provided in managing complex household schedules. Parents’ and children’s enjoyment, curiosity, and sense of adventure encouraged additional travel, transforming routine journeys into playful and memorable family experiences. New ECB travel enhanced family bonding, strengthened intra-household cohesion, and increased children’s willingness to participate in activities that might otherwise have been resisted. Households used ECBs to cultivate sustainable travel identities, model pro-environmental behaviors, and instill active mobility norms in children.</div><div>The paper reframes induced demand and advances research on travel behavior and motivations. It provides valuable insights for policymakers, researchers, and societies, positioning ECBs as a distinct mode in the transition to sustainable mobility.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"43 ","pages":"Article 101217"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145822774","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Rail transit and travel satisfaction: Evidence from a natural experiment in Wuhan","authors":"Zhenhua Li ,&nbsp;Yi Lu ,&nbsp;Jingjing Wang ,&nbsp;Yihao Wu","doi":"10.1016/j.tbs.2025.101211","DOIUrl":"10.1016/j.tbs.2025.101211","url":null,"abstract":"<div><div>The impact of rail transit infrastructure on residents’ travel satisfaction and subjective well-being has gained increasing attention among researchers and policymakers. However, few studies have used longitudinal data to analyze the causal effects of rail transit systems on travel satisfaction and its underlying mechanisms. This study employed a natural experiment approach, using two waves of survey data (2020 &amp; 2021) from 422 participants in Wuhan, China, to assess the effects of a newly opened subway line on travel satisfaction. Applying a mixed-effects difference-in-differences (DID) method, we found that the new subway line significantly improved residents’ travel satisfaction after accounting for socio-demographic and travel attitude covariates. Mediation analysis revealed that this improvement was primarily driven by increased perceived accessibility to downtown and transit stops or stations, as well as a reduction in the number of out-of-home activities on weekends. Heterogeneous analysis indicated that the subway’s benefits are more pronounced among females, individuals under 60 years old, and those from middle-income households. These findings provide new causal evidence on the link between rail transit infrastructure and travel satisfaction, deepening our understanding of this complex relationship and offering practical insights for formulating strategies to improve urban residents’ quality of life.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"43 ","pages":"Article 101211"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145789607","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Meet me halfway – Disentangling the factors affecting leisure joint destination choice","authors":"Joanna Ji ,&nbsp;Benjamin Gramsch-Calvo ,&nbsp;Kay W. Axhausen ,&nbsp;Rolf Moeckel ,&nbsp;Giancarlos Parady","doi":"10.1016/j.tbs.2025.101175","DOIUrl":"10.1016/j.tbs.2025.101175","url":null,"abstract":"<div><div>This study investigates the factors that influence joint-leisure activities and travel between dyads (pairs of friends, family or acquaintances). We draw on a special dataset of self-reported frequently visited leisure destinations conducted in Zurich, and estimate two discrete choice models that consider relationship attributes, such as relationship time length, strength and gender homophily. The first model analyzes home-visits, as the probability of a person hosting a social activity at their place; while the second model is an out-of-home destination choice that quantifies the impact of relationship attributes on the distance traveled for social activities. The findings show that long relationships, (relationship time length <span><math><mo>&gt;</mo></math></span> 7 years), have a higher probability of hosting social activities by 3.87 percentage points, and having a strong relationship (when survey respondent can draw on <span><math><mrow><mo>≥</mo><mn>3</mn></mrow></math></span> expressive resources from the other person) results in higher probability of hosting by 10.5 percentage points. For social activities outside the home, strong ties (<span><math><mrow><mo>≥</mo><mn>3</mn></mrow></math></span> expressive resources from the other person) travel 0.88 km farther on average, and long ties (<span><math><mrow><mo>&gt;</mo><mn>7</mn></mrow></math></span> years) 1.54 km farther, relative to weak (<span><math><mo>&lt;</mo></math></span>3 expressive resources) and short (<span><math><mrow><mo>≤</mo><mn>7</mn></mrow></math></span> years) ties, respectively. However, dyads in a relationship that is both long and strong travel an average of 3.18 kilometers extra than those in relationships that are neither long nor strong, showing that these relationship attributes have an even higher impact when combined.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"43 ","pages":"Article 101175"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145515841","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Changes in travel satisfaction during the day: How pre- and post-trip activities affect travel satisfaction","authors":"Qi Wang ,&nbsp;Na Ta","doi":"10.1016/j.tbs.2025.101193","DOIUrl":"10.1016/j.tbs.2025.101193","url":null,"abstract":"<div><div>Travel satisfaction is a crucial factor in evaluating residents’ subjective perceptions of the transportation system and travel experience, as well as in formulating traffic policies. However, the factors that influence travel satisfaction remain controversial. In previous studies, scholars have focused only on travel behavior and neglected the effects of travel-related activity characteristics. This study aims to address this knowledge gap by exploring the influences of activities before and after trips on trip characteristics and travel satisfaction, using a GPS-facilitated activity diary survey conducted in the Shangdi-Qinghe area of Beijing in 2012. Two structural equation models were constructed to examine the different impacts of activity categories and locations. The results indicate that pre- and post-trip activity characteristics exert both direct and indirect influences on travel satisfaction, with the indirect effects mediated by travel behavior characteristics being more pronounced. Pre-trip recreation activities and post-trip activities at flexible places have direct positive effects on travel satisfaction. Public transit travel and trip duration are important mediators. This study highlights the importance of considering activity-travel interactions in travel-related research.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"43 ","pages":"Article 101193"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145521177","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Long-distance travel impacts of automated vehicles: A survey of American households","authors":"Yantao Huang,&nbsp;Natalia Zuniga-Garcia,&nbsp;Kara M. Kockelman","doi":"10.1016/j.tbs.2025.101168","DOIUrl":"10.1016/j.tbs.2025.101168","url":null,"abstract":"<div><div>This study assesses long-distance (LD) travel demand in near-future scenarios where automated vehicles (AVs) are easily available. Stated and revealed preference data were obtained from 1,004 American adults. The survey includes questions about general LD trip-making behavior with AVs, after investigating the possibility of using AVs to substitute for respondents’ recent LD trips (over 75-miles one-way) prior to the COVID-19 pandemic. After cleaning and weighting the responses, respondents’ willingness to use AVs for a past LD trip are shared, and their future travel behaviors with AVs are explored via statistical models. Ordered logit regression model results suggest AVs would increase the trip-making frequency of high-income households, households with more children, and young people who are part-time employed. The multinomial logit model for overnight stopping suggests that high-income travelers prefer staying overnight in hotels, while young and solo-drivers are expected to more often stay overnight in moving AVs. Mornings departures are preferred for LD travel - both with and without AVs, but some travelers shift to later departures when AVs become available, particularly those who are employed full-time and/or have more children in their household. Results of the mode choice model for mid-range LD travel (200 to 500 miles) suggest those who are unmarried and/or employed full time prefer AVs, while those over age 65 do not. Car and AV cost are the most practically significant variables impacting people’s mode choices, and aircraft and AVs are more appealing than human-driven vehicles for trips over 500 miles long.</div></div>","PeriodicalId":51534,"journal":{"name":"Travel Behaviour and Society","volume":"43 ","pages":"Article 101168"},"PeriodicalIF":5.7,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145759417","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}