Mobility data mining and user modeling
- Geo-location and trajectory mining; activity, behavior, and preference modeling.
- Traffic, demand, and travel-time prediction; mobility forecasting for decisions.
- Large-scale graph and spatiotemporal learning.
AMI’27 · Proposed workshop at WSDM 2027
Agentic Mobility Intelligence
From Mobility Data Mining to Closed-Loop Agentic Systems
Exploring how mobility recommendations, human responses, and agent decisions shape shared transportation systems.
Proposed date19 February 2027
Half-day workshop13:30–17:00 HKT
Attendance modeIn person · Venue TBA
Research focusMining · Recommendation · Interaction
About the Workshop
AMI’27 explores how mobility systems can learn and make decisions when recommendations change user behavior and the conditions for future decisions. Navigation, ride-hailing, and delivery platforms make this challenge concrete: a recommendation that works for one user may have different effects when many users follow it. Understanding these feedback effects requires connecting mobility data mining with recommendation, user modeling, and agentic decision making.
The workshop brings together researchers and practitioners working across these fields, from learning mobility patterns to deploying interactive services. It connects to WSDM’s interests in data mining, recommender systems, and user behavior on large-scale digital platforms. Contributions can address individual components or complete systems, provided they examine the implications for decisions or interaction.
Workshop Objectives
- Connect mining, decisions, and interaction. Develop a shared research agenda linking mobility mining, recommendation, participant behavior, and closed-loop learning.
- Establish when agents add value. Identify evaluation principles that compare agents and conventional recommenders on decision quality, behavioral feedback, and system-wide outcomes.
- Support shared research resources. Explore benchmarks, datasets, simulation environments, and industry collaborations, and share a public synthesis of the workshop.
Topics of interest
We welcome methodological, empirical, systems, dataset, and position contributions across three themes.
Recommendation and agentic systems for mobility
- Platform recommendation, matching, routing, and repositioning.
- Agentic recommender systems; driver, passenger, fleet, delivery, and platform agents.
- Interactive and agentic mobility assistants for contextual, sequential, personalized decisions.
Interaction, feedback, and closed-loop mobility systems
- Human–agent and multi-agent interaction: communication, coordination, and competition.
- Behavioral responses and feedback between recommendations and future mobility states.
- Closed-loop evaluation: simulation, counterfactual analysis, online experiments, and deployment.
Across these themes, we encourage location privacy, equitable access, robustness, user agency, and responsible experimentation.
Who Should Join?
We welcome researchers and practitioners in data mining, recommendation, user modeling, agentic and multi-agent systems, and mobility applications.
The proposed half-day, in-person workshop combines invited talks, contributed talks, posters, and a panel.
Invited speakers and panelists
Invited speakers and panelists will be announced soon.
Speaker informationImportant dates
Paper submission30 November 2026
Author notification18 December 2026
Workshop19 February 2027
Submission and notification deadlines: 23:59 Anywhere on Earth (AoE, UTC−12), following the WSDM 2027 workshop timeline.
Contact the organizing team
For questions about submissions, the workshop or participation, please get in touch.