This is a simple procedure you can follow to create cypehr query for creating Knowledge graph from Your unstuctured text data.
- What is Knowledge graph ? Knowldege graph is simply representing your data in form of graph where node represents entity and edges represents relationships It is very useful in building search system. Also you can implement knowledge graph in RAG to build stronger retriever system.
-
- First clone this repo, Use ipynb file to load you text data and summarize the text data using LLM and langchain.
-
- Use zeroshot.cfg to configure you spacllm for extracting entities and realtionships.
-
- Enter extracted entities and relationships in ChatGpt to get your cypher query.
-
- Create your knowledge graph using that cypher query from ne04j.
I hope you have great time creating your own knowledge graph for better use .