Data Visualization with Gephi

Inner seminar of D&Intel Lab, data visualization with Gephi.

Transcript

1. Southeast University 2/7/2015 1 DATA VISUALIZATION WITH GEPHI Weiwei SUN { “email”: “ww.sun@outlook.com”, “website”: “http://wwsun.me”, “github”: “http://www.github.com/wwsun” }
2. Southeast University 2/7/2015 2 Content • Background • Graph theory and visualization • Gephi introduction • Gephi Demo • Key features of Gephi
3. Southeast University 2/7/2015 3 Facebook, Twitter, Weibo… Platforms like Weibo boost opportunities for connectivity between various types of actors.
4. Southeast University 2/7/2015 4 • At the same time, they produce detailed data traces that are highly centralized and searchable. • Much of these data can be analyzed as graphs
5. Southeast University 2/7/2015 5 Two kinds of mathematics
6. Southeast University 2/7/2015 6 Graph theory http://neo4j-contrib.github.io/asciidoc-slides/content/training/img/konigsberg3.png Leonhard Euler, “seven bridges of kö nigsberg”, 1735 Introducing the “point and line” model
7. Southeast University 2/7/2015 7 New era of graph theory • The network “singularity”: • Computational capacities (memory, speed, interfaces, etc.) • New platforms and datasets • Packaged tools • Different traditions conflate to form network analysis: • Social network analysis and sociometrics • Scientometrics / science and technology studies • Mathematics / physics / computer science • Information and data visualization • Digital sociology / new media studies
8. Southeast University 2/7/2015 8 Two ways of dealing with graphs • Mathematical analysis (graph statistics, structural measures, etc.) • Visualization (network diagram, matrix, arc diagram, etc.)
9. Southeast University 2/7/2015 9 Gephi Introduction • Platform for visualizing and analyzing networks • Using GPU rather than CPU • https://gephi.org Refer: https://gephi.github.io/
10. Southeast University 2/7/2015 10 Gephi in a Nutshell • Field: All types of networks • Types: Directed, undirected, mixed • Size: <1M nodes & edges • Layouts: ForceAltas, Yifan’s Hu Multilevel Helps data analysts to reveal patterns and trends, highlight outliers and tells story with their data.
11. Southeast University 2/7/2015 11 Use Gephi for… • Exploratory Data Analysis • Link Analysis • Social Network Analysis - Demo • Biological Network Analysis
12. Southeast University 2/7/2015 12 Demo • Layout algorithm • Data filtering • Ranking and Grouping • Statics and annotation
13. Southeast University 2/7/2015 13 Layout Layout algorithms give the shape to the graph. - Force-based algorithms - Multi-level algorithms (graph coarsening) Three different force-based layouts of the FB profile: OpenOrd, ForceAltas, Fruchterman-Reingold
14. Southeast University 2/7/2015 14 Metrics • Betweenness, Closeness, Diameter, Clustering Coefficient, Average shortest path, PageRank • Community detection (Modularity) • Random generators https://gephi.github.io/images/screenshots/metricsresults.png
15. Southeast University 2/7/2015 15 Diversity of Network Encoding
16. Southeast University 2/7/2015 16 Software I/O
17. Southeast University 2/7/2015 17 Thank You