Publication Date

3-16-2026

Document Type

Conference Proceeding

Publication Title

Acmlc 2025 Proceedings of 2025 7th Asia Conference on Machine Learning and Computing

DOI

10.1145/3772673.3772682

First Page

10

Last Page

16

Abstract

This project examines how data science and intelligent systems can enhance homelessness prevention efforts by facilitating early identification and optimized service delivery. Using a synthesized dataset reflective of Santa Clara County populations, we developed predictive models to assess risk factors associated with homelessness, such as rent burden, employment instability, eviction notices, and chronic health conditions. These models inform a multi-agent system recommending targeted interventions across medical, food, and job resources. Our system is designed to prioritize individuals most at risk and match them with appropriate services based on availability, proximity, and personal needs. By integrating geographic data and real-time service capacity, the platform improves coordination among job providers, healthcare facilities, and food distribution networks. The goal is to shift the response to homelessness from reactive crisis management to proactive prevention, ensuring that limited resources are allocated equitably and efficiently. This work demonstrates the feasibility of using data-informed strategies to reduce unsheltered homelessness. It provides a scalable framework for policymakers and service agencies to coordinate care, improve outcomes for vulnerable populations, and reduce the long-term societal costs of homelessness.

Keywords

Graph RAG, Knowledge Graphs, Large Language Models, Machine Learning, Multi-agent System

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Department

Interdisciplinary Engineering

Download

Share

COinS