Extracting Cyber Communities Through Patterns

摘要

This paper proposes a new approach to the problem of extracting Web communities. Due to the variety of topics found on the Internet, discovering online communities becomes a really difficult task. In our method, we exploit the observation that pages across the Web refer to other pages of a community using similar text templates in their links. Therefore, by creating patterns that describe these templates we can find a large subset of the pages of a specific community.