Vector search in ArangoDB enables similarity-based queries on unstructured data using embeddings and the FAISS library. The guide demonstrates creating vector indexes, performing approximate nearest neighbor searches with AQL, and combining vector search with graph traversals for advanced use cases like fraud detection. It covers GraphRAG implementation, which merges semantic search with knowledge graphs, and shows how to integrate LangChain for natural language querying. The tutorial includes practical examples of HybridGraphRAG that combines vector search, graph traversal, and full-text search in a single multi-model approach.
Table of contents
What is Vector Search and Why Does it Matter?Setting Up Vector Search in ArangoDBGraphRAG: Combining Vector Search and Knowledge GraphsNatural Language Querying with LangChainWhy Combine Vector Search with Graph?HybridGraphRAG: Combining Vector Search with Graph Traversals and Full-Text SearchHow to Implement HybridGraphRAG in AQLConclusionSort: