[LazyGraphRAG] Questions About LazyGraphRAG's NLP Performance

[SpikeYangRc]

Hi! I'm curious about something in the article from https://www.microsoft.com/en-us/research/blog/lazygraphrag-setting-a-new-standard-for-quality-and-cost/. It mentions that the index building "Uses NLP noun phrase extraction to extract concepts and their co-occurrences". I'm wondering if this NLP approach can achieve the same level of effectiveness as LLM extraction? While I suspect it might be implemented using the NLTK library, I've searched through the related branches and tags but haven't been able to locate LazyGraphRAG's implementation. Would it be possible to point me to the specific branch where this functionality can be found?