CS 432/532 Web Science
Spring 2017
http://phonedude.github.io/cs532-s17/
Assignment #5
Due: 11:59pm March 16
(10 points)
1. We know the result of the Karate Club (Zachary, 1977) split.
Prove or disprove that the result of split could have been predicted
by the weighted graph of social interactions. How well does the
mathematical model represent reality?
Generously document your answer with all supporting equations, code,
graphs, arguments, etc.
Useful sources include:
* Original paper
http://aris.ss.uci.edu/~lin/76.pdf
* Slides
http://www-personal.umich.edu/~ladamic/courses/networks/si614w06/ppt/lecture18.ppt
http://clair.si.umich.edu/si767/papers/Week03/Community/CommunityDetection.pptx
* Code and data
https://networkx.readthedocs.io/en/stable/examples/graph/karate_club.html
http://nbviewer.ipython.org/url/courses.cit.cornell.edu/info6010/resources/11notes.ipynb
http://stackoverflow.com/questions/9471906/what-are-the-differences-between-community-detection-algorithms-in-igraph/9478989#9478989
http://stackoverflow.com/questions/5822265/are-there-implementations-of-algorithms-for-community-detection-in-graphs
http://konect.uni-koblenz.de/networks/ucidata-zachary
http://vlado.fmf.uni-lj.si/pub/networks/data/ucinet/ucidata.htm#zachary
https://snap.stanford.edu/snappy/doc/reference/CommunityGirvanNewman.html
http://igraph.org/python/doc/igraph-pysrc.html#Graph.community_edge_betweenness
(extra credit, 3 points)
2. We know the group split in two different groups. Suppose the
disagreements in the group were more nuanced -- what would the clubs
look like if they split into groups of 3, 4, and 5?