I have implemented the semantic search functionality by creating vector embeddings for all nodes based on their description property using OpenAI api. Then, whenever we search a query, a vector embedding is created for that query, and the 5 most closest nodes are sent as result.
I have implemented the semantic search functionality by creating vector embeddings for all nodes based on their description property using OpenAI api. Then, whenever we search a query, a vector embedding is created for that query, and the 5 most closest nodes are sent as result.