AutoGen Driven Multi Agent Framework for Iterative Crime Data Analysis and Prediction

2025-06-13

AI Summary

This study addresses the challenge of conducting automated, scalable, and iterative analysis and forecasting on crime data under strict privacy constraints. To this end, we propose LUCID-MA, an offline multi-agent framework comprising three tightly coordinated components: an Analysis Assistant, a Feedback Optimizer, and a Prediction Module—enabling spatiotemporal pattern mining, iterative result refinement, and trend forecasting—all executed locally without external API calls. Our key contribution is the first AutoGen-style offline multi-agent system tailored for social science research, featuring a scoring-driven self-assessment mechanism and visual learning trajectory tracking to support up to 100 rounds of autonomous, dialog-based learning. Leveraging LLaMA-2-13B-Chat-GPTQ with domain-specific prompt engineering, we empirically validate high-accuracy pattern recognition and forecasting on real-world crime datasets; iterative refinement significantly improves analytical quality while preserving data privacy, model interpretability, and scalability of autonomous analysis.

Technology Category

• Multiagent Systems: Modeling other Agents
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

This paper introduces LUCID-MA (Learning and Understanding Crime through Dialogue of Multiple Agents), an innovative AI powered framework where multiple AI agents collaboratively analyze and understand crime data. Our system that consists of three core components: an analysis assistant that highlights spatiotemporal crime patterns, a feedback component that reviews and refines analytical results and a prediction component that forecasts future crime trends. With a well-designed prompt and the LLaMA-2-13B-Chat-GPTQ model, it runs completely offline and allows the agents undergo self-improvement through 100 rounds of communication with less human interaction. A scoring function is incorporated to evaluate agent's performance, providing visual plots to track learning progress. This work demonstrates the potential of AutoGen-style agents for autonomous, scalable, and iterative analysis in social science domains maintaining data privacy through offline execution.

Problem

• Analyze spatiotemporal crime patterns autonomously
• Predict future crime trends using AI agents
• Enable offline, privacy-preserving iterative data analysis

Innovation

• Multi-agent framework for crime data analysis
• Offline execution with LLaMA-2-13B-Chat-GPTQ model
• Self-improvement through iterative agent communication