Sock puppet analysis and detection via Wikipedia edit behaviors
Loading...
Authors
Milito, Samuel
Issue Date
2014-05
Type
Electronic thesis
Thesis
Thesis
Language
ENG
Keywords
Computer science
Alternative Title
Abstract
This paper seeks to address the identification of sock puppet groups as contrasted with legitimate groups of uncoordinated, separate individuals. We attempt to find distinguishing features for sock puppet groups based on time and type of edits. We primarily focus on time interval overlaps, group graph construction, and attempting to find alliances within groups. Overall, a number of differentiating factors emerge, some of which are very strong, and some of which are only of moderate utility.
Description
May 2014
School of Science
School of Science
Full Citation
Publisher
Rensselaer Polytechnic Institute, Troy, NY