Agentic Workflow for Travel Data Collection and Behavior Modeling: LLM Chatbot Survey Meets Prediction Models

Forum topic · 小凯 · 2026-08-22

Summary

A paper on arXiv (2608.20320) by Narges Ahmadi, Yubo Jiao, Jonatas Augusto Manzolli, and Jiangbo Yu proposes a three-agent workflow that integrates conversational data collection, structured data processing, and travel behavior prediction. A chatbot-managed, image-enhanced stated preference survey collected mode choices from student commuters under five predefined weather scenarios, yielding 454 subject-scenario observations. Weather associations were analyzed with a multinomial Logit model, while logistic regression and random forests served as machine learning baselines. The authors evaluated nine locally deployed large language models (2B to 35B parameters) across four zero-shot prompting conditions plus extended persona, few-shot, and visual configurations. The random forest achieved 69.6% five-class accuracy, while the best text-only zero-shot LLM reached 69.9% without task-specific fitting. Habitual travel information produced the most consistent gains, expert framing generally outperformed persona-based role-play, and the visual configuration achieved 71.5% accuracy. The work demonstrates that agentic LLM pipelines can unify survey collection and behavioral modeling.

An Agentic Approach for Active Data Collection and Travel Behavior Modeling

Research area: NLP Authors: Narges Ahmadi, Yubo Jiao, Jonatas Augusto Manzolli, Jiangbo Yu Published: 2026-08-22 arXiv: 2608.20320

Overview

Travel behavior research increasingly combines digital data collection and predictive modeling, but these stages are typically developed and evaluated separately. This paper proposes a three-agent workflow that integrates conversational data collection, structured data processing, and behavioral prediction.

A chatbot-managed, image-enhanced stated preference survey collected mode choices from student commuters under five predefined weather scenarios, producing 454 subject-scenario observations. Weather associations were analyzed with a multinomial Logit model, and logistic regression and random forest provided machine learning baselines.

Key findings

Nine locally deployed LLMs (2B to 35B parameters) were evaluated across four zero-shot prompting conditions, plus extended persona, few-shot, and visual configurations.
Random forest achieved 69.6% five-class accuracy.
The best text-only zero-shot LLM reached 69.9% without task-specific fitting.
Habitual travel information produced the most consistent performance gains.
Expert framing generally outperformed role-play personas; the visual configuration achieved 71.5% accuracy.

--- *Automatically collected on 2026-08-22*