Lexical, Vector and Hybrid Search with Elasticsearch
You need to be signed in to add a collection
Search is not just traditional TF/IDF anymore! The current trend of machine learning and models has opened another dimension for search. Generative AI, specifically LLMs, continues to dominate news headlines, marking a turning point in building search applications and finding data. Yet, concerns exist about adapting search solutions to perform semantic searches against our private data. This talk gives an overview of traditional lexical search and its limitations. I’ll explain how vector search can be combined with traditional lexical search to form a hybrid approach using Elasticsearch. Finally, I’ll discuss where LLMs come into play with search and how they can be grounded using RAG.
Transcript
Search is not just traditional TF/IDF anymore! The current trend of machine learning and models has opened another dimension for search. Generative AI, specifically LLMs, continues to dominate news headlines, marking a turning point in building search applications and finding data. Yet, concerns exist about adapting search solutions to perform semantic searches against our private data.
This talk gives an overview of traditional lexical search and its limitations. I’ll explain how vector search can be combined with traditional lexical search to form a hybrid approach using Elasticsearch. Finally, I’ll discuss where LLMs come into play with search and how they can be grounded using RAG.
About the speakers
Developer Advocate and Manager at Elastic