Semantic Search — New Solution to Old Problems

Full-text search is currently undergoing significant changes, with semantic search becoming a hot new trend: From more conversational exploring with LLMs like ChatGPT to enrichment of data and better contextualization.This talk also looks at the new foundation with vector search and limitations of our new tools.This talk starts with the classic TF/IDF search that has been in use for decades and then shows what semantic or vector search can add. You get an overview of the ML/model side as well as a dive into the common search approaches with KNN and ANN. And we discuss where hybrid search of both old and new models comes into play.

Target Audience: Developer
Prerequisites: Database experience and interest in machine learning
Level: Advanced 

Philipp lebt für technische Vorträge und Demos. Nachdem er mehr als zehn Jahre als Web-, Infrastruktur- und Datenbank-Entwickler gearbeitet hat, ist er mittlerweile Developer Advocate Manager bei Elastic – dem Unternehmen hinter dem Elastic Stack, bestehend aus Elasticsearch, Kibana, Beats und Logstash.

Philipp Krenn
Track: Track 1
16:00 - 16:45
Vortrag: Di1.4

Vortrag Teilen