YoVDO

Personalizing Search Using Multimodal Latent Behavioral Embeddings

Offered By: OpenSource Connections via YouTube

Tags

Semantic Search Courses Vector Databases Courses Collaborative Filtering Courses Retrieval Augmented Generation (RAG) Courses

Course Description

Overview

Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore the cutting-edge approach to personalizing search using multimodal latent behavioral embeddings in this 45-minute conference talk from Haystack US 2024. Delve into the importance of incorporating user context and behavioral signals to optimize search relevance, moving beyond traditional keyword-based and content embedding methods. Learn how to integrate user behavior into modern search retrieval pipelines for RAG and end-user search, combining content, domain, and user understanding for a holistic approach to search relevance. Discover techniques for training embedding models using behavioral signals, implementing personalized search experiences, and applying appropriate contextual guardrails. Gain insights into traditional signals-based models for AI-powered search and their mapping into multimodal embedding approaches. Witness live, open-source code examples demonstrating how modern hybrid search approaches can learn user and group affinities. Benefit from the expertise of Trey Grainger, lead author of "AI-Powered Search" and founder of Searchkernel, as he shares his extensive experience in developing semantic search, personalization, and recommendation systems.

Syllabus

Haystack US 2024 - Trey Grainger: Personalizing search using multimodal latent behavioral embeddings


Taught by

OpenSource Connections

Related Courses

Vector Similarity Search
Data Science Dojo via YouTube
Supercharging Semantic Search with Pinecone and Cohere
Pinecone via YouTube
Search Like You Mean It - Semantic Search with NLP and a Vector Database
Pinecone via YouTube
The Rise of Vector Data
Pinecone via YouTube
NER Powered Semantic Search in Python
James Briggs via YouTube