On Determining Communities in the Web

One of the characteristic features of the World Wide Web is that it allows anyone to publish web pages, and to link their pages to any other web pages. The emergent result of this process is that communities arise that are focused around subjects of common interest. Identifying these communities can be challenging, due to the large size of the web graph and the unstructured nature of publishing web pages. In addition, traditional notions of community are based upon community members being in physically adjacent spaces, but geographic proximity is not present on the Web. In this paper, we discuss recent methods of detecting communities on the Web, and propose new directions.