Georgia on my Mind: NLP Meets Social Network Analysis for Exploring New Domains
summary
Method combines social network analysis and NLP to turn a few English Wikipedia seed pages into usable maps of a domain using networkx spaCy wikipedia. The pipeline makes link graphs, runs PageRank and betweenness, peels k‑cores (five‑core revealed two clear components), detects communities with Len, and builds per‑sentence entity co‑occurrence graphs so clusters actually separate Tbilisi neighborhoods and institutions. Results move from a useless “hairball” to specific surprising findings (odd PageRank hits, clearer betweenness, meaningful k‑core split) and a working entity network that yields actionable topics.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:55Networks and text
- 6:58Building network from Wikipedia
- 10:21Important nodes in network
- 17:39Network structure with K-core
- 22:26NLP deep dive into nodes
- 28:28Insights and applications
- 29:59Q&A on embeddings
related talks
01h 41mGraph Intelligence: Enhance Reasoning and Retrieval Using Graph Analytics
30mFast and scalable domain-specific knowledge graphs generation
41mManaging Connected Big Data in Art with Neo4j Graph Database
28mA new wave of Network Science - from fantasy to applications
01h 38mNetwork Science for Fun and Profit
41mLarge Road-Network analysis with GraphX
01h 32mConnected: A Social Network Analysis Tutorial with NetworkX
01h 40mKnowledge Graphs & GraphRAG: Techniques for Building Effective GenAI Applications