Duke ProfMatch: AI-Powered Research Collaboration Tool

Feb 18, 2024
Repo
Photo by rawpixel on Unsplash

Duke ProfMatch is an AI-powered platform designed to connect students with professors whose research aligns with their interests. Inspired by the difficulty students face in identifying potential collaborators, the tool uses natural language queries to return curated lists of professors along with details on their research, publications, and contact information.

Built with Graph Retrieval-Augmented Generation (Graph RAG), Neo4j, and OpenAI models, ProfMatch features a visual graph of Duke’s academic network, allowing users to explore connections across departments, schools, and research areas. The system was created by transforming Scholars@Duke data into a graph database, integrating prompt-engineered retrieval pipelines, and deploying a Streamlit frontend.

ProfMatch makes academic networking more accessible and engaging, with future plans to expand to other universities and add advanced features such as keyword alerts, collaboration suggestions, and improved AI-driven matching.

Graph RAG Academic Networking Neo4j OpenAI Streamlit LLM
Chao Péter Yang
Chao Péter Yang
ML Research Assistant

My research focuses on symbolic music generation and graph machine learning, with broader interests in generative modeling, agentic AI systems, and large models. I aim to bridge theory and practice in data science to create both scientific and real-world impact.