Demystifying Social Network Analysis (SNA)
summary
Five case studies show concrete claims and tools not platitudes: data collection uses Python and Twitter APIs with 30–40 million tweets for the 2019 election, backups and redundancy after a one‑drive crash, plus IMDb scraping for Bollywood. Analyses combine NLP (word embeddings, sentiment), network metrics like eigenvector centrality and betweenness, visualization and computer vision for memes. Surprising results: Bollywood top‑30 form a tight clique with average degree ≈27–28, MeToo centers on four influencers, legal networks show low clustering, and trolling and image‑based manipulation rise sharply.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:23Case Studies
- 3:20Indian Politics 2014
- 11:01Network Analysis
- 15:29Insights
- 22:04Reply Network
- 26:14Connection
- 28:13Clustering
- 29:08Content is King
- 31:06Eigenvector centrality
- 32:17Other dimensions
- 32:52Final graph
- 33:16Legal Network
- 33:51Prashant Bhushan
- 34:46Outcome
- 35:07Questions
related talks
28mSocial n/w Analytics to enhance Marketing Outcomes in Telecom Sector
01h 27mSocial Network Analysis with Python
24mHands-On Network Science
01h 38mNetwork Science for Fun and Profit
28mA new wave of Network Science - from fantasy to applications
26mSocial Networks and Protest Participation: Evidence from 130 Million Twitter Users
43mVisualising networks
01h 32mConnected: A Social Network Analysis Tutorial with NetworkX