Jul 2026· Transportation Research Record· 0 citations· 56 references
TL;DR
This study proposes an integrated analytical process that combines revealed-preference survey data, route-level attributes derived from a digital trip planner, machine-learning classifiers, and explainable artificial intelligence (XAI) methods to evaluate predictive performance and behavioral interpretation jointly.
Abstract
Modeling mode choice is essential for designing efficient and sustainable mobility systems. Revealed-preference surveys provide valuable information, but they rely on self-reported data, which can be biased and are typically unavailable for unchosen alternatives. This study proposes an integrated analytical process that combines revealed-preference survey data, route-level attributes derived from a digital trip planner, machine-learning classifiers, and explainable artificial intelligence (XAI) methods to evaluate predictive performance and behavioral interpretation jointly. Using a dataset of 1,372 trips collected in a university commuting context as an illustrative application, survey responses were enriched with mode-specific travel times and geometric characteristics of planner-recommended routes obtained from Google Maps’ application programming interface. Four tree-based classifiers were evaluated in a leak-free validation framework, and model behavior was interpreted using SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations) at both the global and local levels. The results indicate that when route-level geometric attributes were combined with revealed-preference survey data, predictive accuracy remained comparable, whereas interpretability improved substantially. XAI analyses revealed that route characteristics such as straightness, sinuosity, and angular deviation emerged as significant predictors that modulated perceived travel effort, particularly for walking and public transport, despite their limited impact on aggregate performance.
Overall, the study shows how conversational surveys, structured data processing, conventional behavioral modeling, machine learning, and multimodal LLM prediction can be coordinated within an auditable multi-agent workflow.
N. Ahmadi, Yubo Jiao, J. Manzolli et al.· 0 citations
A tourist may describe a desired day as 'relaxed, coastal, suitable for parents, and not too crowded,' whereas a route optimizer requires numerical attributes and explicit constraints. This paper connects these two representations without asking a large language model (LLM) to draw the route itself. The LLM parses a natural-language request into a preference profile; a semantic-spatial network then links that profile to attraction attributes, travel connections, visit durations, and congestion information. Route selection is performed by a multi-criteria model that evaluates preference fit together with distance and time costs. The framework is examined using six attractions in Dalian and four traveler profiles. Compared with the shortest-path baseline, the LLM-assisted method increases the reported preference-matching degree by about 29.1%, although it does not always return the minimum-distance itinerary. The result suggests a practical division of labor: language modeling handles ambiguous user intent, while an explicit optimizer remains responsible for spatial feasibility and resource limits.
Kai-Nan Ma, Wei Pan· Frontiers in Computing and I...· 0 citations
Current pedestrian routing assumes all pedestrians are identical and optimizes a single objective such as journey time, despite 6 decades of evidence showing personality consistently influences route preference. This paper develops a route, an agent-based framework that maps Big-Five-plus-sensation-seeking profiles to eight route attributes through a multinomial logit (MNL) cost function governed by an 8 × 6 personality, attribute interaction matrix (γ). The framework integrates a Mesa 3.x simulation on the Boğaziçi University campus map (740 nodes, 1,892 edges, SRTM elevation 4–135 m), a five-layer multi-modal pathfinder governed by a personality–mode matrix (δ), and a p-median facility-location optimiser that converts personality-segmented demand into infrastructure placement. Three converging lines of evidence support the approach. A 2,000-agent simulation produced significant alignment improvements on all nine pre-registered tests at p < .001 (Reserved archetype Cohen's d = 1.563), and was robust to terrain. A real Phase 1 screening survey (n = 26, 19 eligible) confirmed 16 of 23 hypothesised δ sign directions (70 %, binomial p = .047). A synthetic pipeline validation produced large subjective-measure effects (satisfaction d = 1.169, commercial viability d = 1.445) while objective deviation stayed flat. A multi-modal extension cut travel time by 43 %; the optimiser recommended 23 scooter docks, saving 2,397 person-hours per day at Gini = 0.007. Personality matters for routing, as does the technology that underlies it.
C. Kenter, Ilgin Gokasar· Journal of Expert Systems an...· 0 citations
There are significant social, environmental, and public health advantages to encouraging cycling over automobile transportation. Although the "x-minute city" concept aims to decentralize urban services within a short bike ride or walk, there is currently little empirical modeling of how this framework affects individual behavioral choices. This study assesses the effect of proximity on bicycle adoption using intelligent urban modeling. We generated Multinomial Logit (MNL) models to examine mode and destination choice dynamics using two stated preference experiments conducted through a face-to-face survey (N=390) in Kerman, Iran. Travelers are successfully diverted from the city center by creating decentralized, replicated urban services within a 10-minute cycling radius, according to the predictive analytics. When combined with push-factors like congestion pricing and smart infrastructure expenditures like dedicated parking and bike lanes, this behavioral shift is further enhanced. For transportation planners and legislators using data intelligence to create sustainable, polycentric smart cities, these results offer crucial quantitative insights.
Hamidreza Fateh, Saeid Sherafatipour, M. Saffarzadeh· Journal of Urban Intelligenc...· 0 citations
The 15-minute city promotes access to everyday services within a short walk or bicycle ride, but its relationship with observed mobility remains difficult to quantify. We investigate this relationship in the Paris metropolitan area using mobility trajectories from the NetMob 2025 Data Challenge, enriched with INSEE sociodemographic data and OpenStreetMap points of interest (POIs), yielding approximately 70,000 trip segments after stop-based segmentation and data cleaning. We construct walking- and cycling-based indicators of local service availability and examine their associations with trip duration, transport mode, and short-trip car use. Higher POI availability is associated with less private motorized travel and more active mobility, although this relationship is substantially weaker in the outer agglomeration. Gradient-boosted tree models interpreted with explainable machine-learning methods consistently identify trip purpose, home--work distance, local service availability, vehicle ownership, public-transport subscription, and sociodemographic context as important predictors. For short trips, high POI density is associated with lower car use, while car ownership and driving-licence availability are associated with higher predicted car use; where services are sparse, public-transport subscription is associated with lower predicted car dependence. Finally, explainable AI (XAI) methods are used to examine how feature attributions change under alternative assumed variable orderings. The results are consistent with central assumptions of the 15-minute city while revealing substantial spatial and demographic heterogeneity. They also demonstrate how explainable machine-learning methods can complement accessibility indicators and identify locally relevant hypotheses for urban-mobility policy.
A. Molnár, Csaba I. Sidló, Rita Rónai et al.· 0 citations
A spatial proximity weighted eXtreme gradient boosting (XGBoost) ensemble to predict daily metro ridership under holidays, weekdays, and weekends is proposed, providing strong generalizability and practical value for urban transit planning and demand forecasting.
Xinyu Hu, Cong Qi, Yunpeng Zhao et al.· Journal of Transportation En...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
