do you have user manual or paper how to use it and theory behind pls

[Sandy4321]

great code but how to use , do you have user manual or paper how to use it and theory behind pls

[rangehow]

Paper：https://arxiv.org/abs/2404.16130 After reading the paper, you can refer to many examples in the example folder under the repository.

[Sandy4321]

but this paper is not about nanograph, but about Microsoft graphrag. Microsoft graphrag is really bad, it is why people looking for new approaches like lightgraph. It is why your code become popular. so do you have description what you doing in code?

[rangehow]

Compared to the differences between lightrag and graphrag, our approach is algorithmically almost identical to graphrag (except for some subprocesses we've omitted, as detailed in the readme). Lightrag, on the other hand, is an incremental effort that differs from graphrag at the algorithmic level. The popularity of our repository stems from our commitment to clarity in implementing the algorithms, avoiding overly complex abstraction layers. This makes it easy for users to modify, which is why many subsequent modifications based on graphrag have leveraged our code.

By reading the original GraphRAG paper, you can clarify the meaning behind many variables and abstraction layers in the project. Then you can look through the examples to quickly adapt it to your task.

[Sandy4321]

but you use grpg neural network to get nodes embeddings. But it is slow and embeddings are bad? did you try other approaches to find nodes similarities ?