Catalog & Cocktails: The Honest, No-BS Data Podcast Understanding the Role of Knowledge Graphs on Large Language Model’s Accuracy for Question Answering on Enterprise SQL Databases
14 snips
Nov 22, 2023 Experts Juan Sequeda, Dean Allemang, and Bryon Jacob discuss the impact of knowledge graphs on language models' accuracy for question answering on enterprise SQL databases. They highlight the benchmark results showing significant improvement in accuracy and emphasize the importance of investing in knowledge graphs to enhance AI systems.
AI Snips
Chapters
Transcript
Episode notes
Enterprise Data Complexity
- LLMs have accelerated text-to-SQL, but current examples are often simple.
- Enterprise data and questions are more complex, requiring a different approach.
Knowledge Graph Hypothesis
- Knowledge graphs improve LLM accuracy on enterprise data questions.
- This research compares LLM performance with and without knowledge graphs.
LinkedIn Popularity
- Tim Gaspur highlights the practical implications of the research.
- Many want to use LLMs with internal data, but question its effectiveness.