Authors

Xinyu Gong, Feiyang Ren, Sridevi Turaga

Research Question

Can a chatbot moderate social interactions focusing on urban issues, and how can a framework for its usage be developed to recommend urban management policies?

Background

This project explores the potential of AI-powered chatbots to facilitate structured, multi-user dialogue around complex urban issues. It examines how conversational agents can be designed to support civic interaction, mediate conflicts, and generate actionable insights for urban governance. The study integrates perspectives from human-computer interaction, conflict resolution theory, and AI system design to assess how digital tools promote inclusive participation in policy discussions. By reviewing existing chatbot systems, the project critically examines intervention frameworks and the role of chatbots in mediating public discourse. The aim is to develop a unified framework for designing intelligent systems for civic dialogue and community feedback.

Methodology

The methodology is divided into experimental design and chatbot development. Experimental design involves internal testing, defining interaction protocols for group chat sessions, and establishing a systematic framework for data collection and analysis. This includes logging conversation flows, tracking user intent, and assessing engagement patterns under controlled conditions. Chatbot development involves designing a multi-user chatbot system with algorithmic flowcharts and technical specifications, and developing a multi-agent simulator. The simulator enables AI agents to interact on assigned topics, applying conflict resolution theories to detect, analyze, and resolve conflicts. Testing involves manual annotation of conflict labels comparing results with mediator response.

Deliverables

• Technical Report
• Multi-User Chatbot
• Multi-Agent Chat Simulator
• Systematic AI Review

Datasets