Knowledge data management is the ability to search and manage data based on semantic relationships using vectors and graphs. It can be used to retrieve and manage knowledge data provided to the Large Language Model (LLM) in systems that use the method of Search Extension Generation (RAG).

In addition to the existing search methods for knowledge data stored in FEP, FEP can use the Knowledge Data Management feature to search for knowledge data in three ways:

• Vector data management feature

• Text semantic search and automatic vectorization
  Hybrid search is also available, combining semantic and full-text search

• Graph Management feature

This section describes how to set up each feature.