How to build an advanced semantic search engine with hybrid search | Elasticsearch Coding Sessions
Get ready to say 'Hasta la vista, baby' to outdated search methods as we take a closer look at semantic search, using a data set of some all-time favorite sci-fi 🤖 and horror movies!
Join Ugo Sangiorgi, principal product marketing engineer, for a 20-minute coding session to learn about:
- The fundamentals of full-text search and its token-matching process
- How semantic search leverages vector embeddings for deeper understanding
- Step-by-step implementation of hybrid search using Elasticsearch
- The role of retrievers such as reciprocal rank fusion (RRF) and linear retrievers in merging search methods
💡 Key Highlights:
- Live demonstration using a movie search application featuring some classic sci-fi and horror films
- Detailed walkthrough of configuring search fields and connectors
- Insights into scoring mechanisms and balancing semantic with full-text search
0:00 - Introduction
0:55 - Lexical search
06:38 - Semantic search
12:00 - Hybrid search
17:18 - Q&A
📚 Resources:
- GitHub Repo: Source code and materials from the session: https://github.com/ugosan/elastic-coding-sessions
- Search Labs: Elastic's blog for developers and data scientists: https://www.elastic.co/search-labs
- Follow Ugo on LinkedIn https://www.linkedin.com/in/ugosangiorgi/ or catch up on his latest Search Labs blogs: https://www.elastic.co/search-labs/author/ugo-sangiorgi
If you’re looking to add AI-driven search to your app, product, or website, this session is for you. Engage with us in the chat, share your thoughts, and feel free to ask questions. Let's dive into the world of hybrid search with Elasticsearch!
👍 Like, Share, and Subscribe for more Elasticsearch Coding Sessions!