Introduction to NLP models and vector search: Part II

Tom Grabowski

Principal Product Manager

Elastic

Nick Chow

Prinicipal Product Manager

Elastic

Gilad Gal

Principal Product Manager I

Elastic

This is an update of our Introduction to NLP: Part I session which includes updates from 8.1 - 8.3 of the Elastic platform.

Introducing modern NLP and native vector search in Elasticsearch. Leverage new ML models to understand context, increase speed and improve results. Unlock even more advanced text analytics like semantic sentence embedding and Question Answering NLP PyTorch models with significantly less effort and time. Get the latest updates on vector search including; filtering, radius queries, incremental indexing, and more. Start with pre-built models or scale your own.

Highlights:

• Using dense vector fields in Elasticsearch for vector similarity
• Filtering a vector search
• Using a radius query to define the subset of results that is deemed relevant to the query
• Handling incremental changes to the index
• Building applications using semantic search with NLP models
• Working with HuggingFace PyTorch models
• Using vectors and NLP to create modern semantic search applications

Additional resources:

• Get your search tool kit for the AI era — the Elasticsearch Relevance Engine (ESRE)
• Introduction to NLP models and vector search: Part I
• Introduction to modern NLP with PyTorch in Elasticsearch
• Introducing Approximate Nearest Neighbor (ANN) search
• NLP getting started and end-to-end examples (blog series)
• Documentation: NLP
• Documentation: Dense Vector Field Types
• Want to try it for yourself? Learn more about Elastic Cloud or, if you're ready to get started, spin up a free 14-day trial