Detecting and Understanding Web communities

Web communities comprising of documents and/or users activities have been formed and are continuously expanding and transforming as Web users role shifts from typical navigations to content managing and regulating. Defining, deriving and exploiting communities is not a trivial task since several parameters (large-scale, complexity, evolving information etc) are involved. This paper aims at providing answers for crucial questions raised about communities emerging in the Web and it summarizes different community definitions such that then, the problem of community detection (which is well matured and researched in the past) is understood. The paper emphasizes and discusses the most important methodologies and techniques which deal with large populations of Web documents participating in vast hyperlinked networks, or networks formed from crawling (part of) the web and more recently, networks reflecting the social relations and/or interactions among people. It is important to understand and categorize community identification efforts by taking into account that different levels of granularity and different views ar