Overview
Explore the new era of search experience with vector search in this 27-minute ElasticON EMEA conference talk. Discover how vector search enables queries without specific keywords and applies semantic search to text, images, and audio. Learn about intelligent discovery methods like recommendation engines in e-commerce and question-answering using transformer models. Understand Elastic's support for vector search and NLP since the 8.0 release, including approximate nearest neighbor search and importing pretrained PyTorch models. Gain insights on combining traditional retrieval techniques with vector search to achieve optimal results. Explore use cases, conceptual overviews, and a demo showcasing textual search, OpenAI Clip, model uploading, and image embedding.
Syllabus
Introduction
Why use Vector Search
Text Matching
Intent Matching
Keyword Search
Garment Search
Conceptual Overview
Use Cases
Vector Search
Machine Learning Model
Model Hub
Search
Demo
Textual Search
OpenAI Clip
Uploading Models
Embedding Images
Results
Development
Summary
Taught by
Elastic