AI-Semantic Search¶
AI-Semantic Search is an advanced search powered by Artificial Intelligence (AI) that understands the meaning and context of search queries and documents.
Overview¶
AI-Semantic search enables users to find information using natural language questions and understands the semantic meaning of content. It's particularly effective when users don't know the exact keywords or when the relevant content uses different terminology than the search query.
Recommended Search Type
While AI-Semantic search is powerful for natural language queries, we recommend starting with Hybrid Search as it combines the benefits of both semantic and lexical search. Use pure AI-Semantic search when you specifically want to prioritize natural language understanding over keyword matching.
Key Features¶
- Natural Language Understanding: Users can search using complete questions or conversational phrases
- Contextual Understanding: Finds relevant content even when keywords don't exactly match
- Intent Recognition: Understands the underlying purpose of the search query
- Automatic Typo Handling: Built-in tolerance for spelling mistakes and variations
Best Use Cases¶
AI-Semantic search excels in scenarios where:
- Users ask complex questions in natural language
- Content meaning is more important than exact keyword matches
- Search needs to understand synonyms and related concepts
- Documentation or content uses varying terminology for similar concepts
Why Choose AI-Semantic Search¶
Choose AI-Semantic search when you need:
- Superior natural language understanding
- Context-aware search results
- Flexibility in how users can phrase their queries
- Search that works well with conversational queries
Search Type Selection
You can easily switch between search types by changing the search_type
parameter in the Search Documents API. Try hybrid
first, then switch to ai_semantic
if you need stronger emphasis on natural language understanding.
Retrieval-Augmented Generation (RAG)
If you need AI-generated answers instead of document search results, check out our RAG capability.