Internet identity analysis and similarities detection

Abstract

Growing popularity of Web 2.0 systems created huge set of publicly available data, which is continuously expanded by users of the Internet. The anonymity of publications in Web systems encourages some users to publish false or illegal statements. Tools for identifying portal users, who publish such posts could result in higher quality of information and could be useful for law enforcement services. In this paper a method for finding similar Internet identities is introduced. Detected similarities can be used for finding several accounts of the same person. The method is based on calculating various measures characterizing forums users. It uses Web crawling system to collect data from forums. A prototype system for finding similar users is described and tests results are presented. © 2012 Springer-Verlag.