Loading...
Please wait, while we are loading the content...
Similar Documents
Clustering of Wikipedia Pages on Edit Behaviors
| Content Provider | Semantic Scholar |
|---|---|
| Author | Talukder, Nilothpal Li, Haiqiong Magdon-Ismail, Malik |
| Copyright Year | 2011 |
| Abstract | We consider the edit history of Wikipedia to perform clustering of the pages. We conjecture that the editors exhibit homophily or high correlation (in terms of the topics of interests). Therefore, it is possible to utilize the edit history to cluster pages having same or closely related topics. We validate our clustering results with the list of categories and the incoming and outgoing links on the Wikipedia pages. We use k-means to perform page clustering. Typically, Wikipedia page editors demonstrate multiple interests and Wikipedia pages list multiple categories, whereas k-means delivers only partitioning. Therefore, we also study the results from a clustering algorithm called “Connected Iterative Scan” that produces overlapping communities from a social graph. We also study page dynamics which can potentially be incorporated into the clustering of pages. Keywords-Wikipedia page edits, collaboration graph, clustering |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.cs.rpi.edu/~magdon/courses/casp/projects/TalukderLi.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
