// MODEL OPTIMIZATION AND PROMPT SYNTAX TERM
Hybrid Search
Hybrid search combines different search techniques, such as keyword matching and semantic understanding, to get more comprehensive and accurate results. It leverages the strengths of multiple approaches.
TECHNICAL DEFINITION
A search paradigm that integrates multiple retrieval methods, typically combining lexical (e.g., BM25, keyword) and semantic (e.g., dense vector embeddings) approaches to leverage their respective strengths and improve overall recall and precision in information retrieval.
BACKGROUND
A large language model (LLM) is a neural network trained on a vast amount of text for natural language processing tasks, especially language generation. LLMs can typically generate, summarize, translate, and analyze text in many contexts, and are a foundational technology behind modern chatbots. Biased or inaccurate training data can make an LLM's output less reliable.
READ MORE ON WIKIPEDIASYNONYMS & ALIASES
- Multi-modal search
- Blended search
- Combined search
- Dual-mode retrieval
- Composite search
USAGE NOTE
Hybrid search is increasingly popular in RAG systems to overcome the limitations of purely keyword-based or purely vector-based retrieval.
DEVELOPERS
Organizations developing technology related to Hybrid Search.
Develops Elasticsearch, a distributed search and analytics engine that supports advanced keyword search and has integrated vector search and hybrid search capabilities, crucial for RAG and AI applications.
An open-source vector database that natively supports hybrid search, combining vector similarity search with keyword search to improve retrieval for AI-powered applications like RAG.
A managed vector database service that provides hybrid search functionality, allowing users to combine semantic search (vector embeddings) with traditional keyword filtering for more relevant results.
A high-performance vector similarity search engine that offers native support for hybrid search, combining vector search with payload filtering and keyword matching to enhance retrieval accuracy.
The company behind Milvus, an open-source vector database, and offers Zilliz Cloud. They provide solutions for vector search that enable hybrid search by combining vector similarity with scalar filtering and advanced query capabilities.
An open-source serving engine for large-scale AI applications, developed by Yahoo!, that is renowned for its powerful hybrid search capabilities, combining vector-based recommendations with lexical search features.
A cloud search service offering native hybrid search capabilities that merge vector search with traditional keyword and filter-based search, enabling more comprehensive and relevant information retrieval for AI applications.
An open-source search and analytics suite, forked from Elasticsearch, that is actively developing and implementing hybrid search features to combine vector search with lexical search for improved relevance in AI-driven retrieval.