Leskovec www 2010

From Cohen Courses
Revision as of 22:14, 16 February 2011 by Subhodee (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

Citation

Leskovec, J., K. J Lang, and M. Mahoney. 2010. Empirical comparison of algorithms for network community detection. In Proceedings of the 19th international conference on World wide web, 631–640.

Online version

WWW 2010

Summary

This paper presents an empirical evaluation of algorithms for the problem of Network Community Detection. They use 40 different real world networks, compare 12 objective functions and 8 approximation algorithms and report their findings.

  • They find that network detection for large graphs can be surprisingly intricate
  • They find that most methods exhibit qualitatively similar behavior
  • However depending on the objective certain kinds of clusters tend to be preferred which the authors refer to as systematic biases of the method.
  • The authors suggest that a regularization like approach would help correct for the specific deficiencies of the methods.