The GraphRAG Manifesto: Unlock Better GenAI Results With Knowledge Graphs | Read Now

Neo4j logo

Nodes2024

Dev Conference by Neo4j

Register for NODES 24

You only need to register once to attend all sessions.

Unified Learning Approach using KG-RAG-Chatbot (Ukit)

Session Track: AI

Session Time:

Session description

Though we live in an interconnected world, highly correlated concepts are taught in isolation leading to disconnected learning. Ukit aims to solve this by equipping students with an integrated knowledge map. Their goal is to establish connections between concepts across various subjects and illustrate their logical convergence using HuggingFace and Neo4j graph and vector databases. You will learn about Knowledge Graph (KG) construction, language models fine-tuning using few-shot learning, and lower-rank adapters to refine entity categorization within the KG framework. This focuses on assigning neighborhood weights and minimizing redundancy. They are eager to share their learnings with the machine learning and research community.

Speaker

photo of Vidya S

Vidya S

Data Scientist, amii

Vidya shree S is a Machine Learning Resident on the Advanced Tech team of Amii. She brings in five years of experience in machine learning, predictive modelling, and end-to-end data management across diverse industries including supply chain, manufacturing, and e-commerce. She completed her MEng at the University of Waterloo. Her research interests include sequential data, feature engineering and deep learning. Outside of work, Vidya enjoys reading books, hiking and music.