Promoting TunkRank


Early last year, Daniel Tunkelang proposed a way to measure people’s influence on Twitter; this metric was dubbed TunkRank, and Jason Adams put up an implementation of it that people could use to calculate their (and others’) scores. The site has been evolving, and getting slicker. It even has an API for incorporating these scores into other applications.

The basic premise of the algorithm is that its not how many followers you have, but how influential they are. Your influence flows from them. For those interested in more details and rationale about algorithm, Daniel’s slides from a recent talk offer a nice overview. What’s also interesting, as pointed out in the comments on his post, is that this model, proposed on the blog and never published in a peer-reviewed forum, has become quite influential.

Share on: 


  1. […] This post was mentioned on Twitter by Jason Adams, Gene Golovchinsky and 윤효석 (Hyoseok Yoon), TunkRank. TunkRank said: RT @HCIR_GeneG: Posted "Promoting TunkRank" #twitter […]

  2. At CERI2010, Daniel Gayo-Avello presented an analysis of PageRank, HITS, NodeRanking, TunkRank, TwitterRank, concluding that TunkRank was best on a large twitter corpus. See or a summary of the CERI2010 paper with links to various related stuff .

  3. Nice that they cited the blog!

Comments are closed.