An Agentic Approach for Active Data Collection and Travel Behavior Modeling

Forum topic · 小凯 · 2026-08-24

Summary

Researchers propose a three-agent workflow that unifies conversational data collection, structured data processing, and behavioral prediction for travel behavior research. A chatbot-managed, image-enhanced stated preference survey collected students' mode choices across five predefined weather scenarios, yielding 454 respondent-scenario observations. A processing agent then cleaned and structured the data, which a prediction agent used to train weather-sensitive travel demand models. The integrated approach improved data quality and model interpretability, while the agentic architecture enables modular updates and reuse across studies. Published on arXiv (2608.20320) by Narges Ahmadi, Yubo Jiao, Jonatas Augusto Manzolli, Jiangbo Yu, and Luis Miranda-Moreno, the work offers a blueprint for combining active data collection with behavioral modeling, with implications for adaptive transportation systems and real-time demand management.

Paper Overview

Field: NLP Authors: Narges Ahmadi, Yubo Jiao, Jonatas Augusto Manzolli, Jiangbo Yu, Luis Miranda-Moreno Published: 2026-08-22 arXiv: 2608.20320

Abstract

Travel behavior research increasingly combines digital data collection with predictive modeling, yet these stages are often developed and evaluated separately. This study proposes a three-agent workflow integrating conversational data collection, structured data processing, and behavioral prediction.

Key Points

Three-agent workflow: The pipeline integrates (1) conversational data collection, (2) structured data processing, and (3) behavioral prediction as modular agents.
Survey design: A chatbot-managed, image-enhanced stated preference survey captured students' mode choices under five predefined weather scenarios, producing 454 respondent-scenario observations.
Processing and prediction: A processing agent cleaned and structured the collected data; a prediction agent then trained weather-sensitive travel demand models.
Results: The integrated approach improved data quality and model interpretability compared to treating collection and modeling separately.
Modularity: The agentic architecture allows modular updates and cross-study reuse.

Significance

The work provides a blueprint for combining active data collection with behavioral modeling, with implications for adaptive transportation systems and real-time demand management.