Explainable and Responsible AI with GNNs and GraphRAG

Speaker:  Arijit Khan – Bowling Green, USA
Topic(s):  Artificial Intelligence, Machine Learning, Computer Vision, Natural language processing , Information Systems, Search, Information Retrieval, Database Systems, Data Mining, Data Science

Abstract

Graph data--whether social or biological networks, financial transactions, or knowledge graphs--models entities as nodes and their relationships as edges. Deep learning and generative AI techniques--like graph neural networks (GNNs) and large language models (LLMs)--have become ubiquitous in domains ranging from cheminformatics and bioinformatics to fraud detection, question answering, and recommendation. Yet these models often act as black boxes, and LLMs in particular can hallucinate due to outdated or inconsistent knowledge. In the first half of this talk, I’ll show how data science and systems approaches can produce user-friendly, configurable, queryable, and robust explanations for GNNs. In the second half, I’ll describe our latest work on knowledge graph–based, retrieval-augmented generation methods (GraphRAG) that enhance LLM robustness and performance across diverse data science applications.

About this Lecture

Number of Slides:  45
Duration:  60 minutes
Languages Available:  English
Last Updated:  18/02/2026

Request this Lecture

To request this particular lecture, please complete this online form.

Request a Tour

To request a tour with this speaker, please complete this online form.

All requests will be sent to ACM headquarters for review.