Traditional link-based schemes for identification of web communities focus on partitioning the web graph more sophisticatedly, without concerning the topical information inherently held by web pages. In this paper, we give a novel method of measuring the topicality of a hyperlink according to its context. Based on this, we propose a topical maxflow-mincut algorithm which incorporates topical information into the traditional maxflow-mincut algorithm. Experiments show that our algorithm outperforms the traditional algorithm in identifying high-quality web communities. © Springer-Verlag Berlin Heidelberg 2009.
CITATION STYLE
Miao, Y., & Li, C. (2009). Topical analysis for identification of web communities. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5866 LNAI, pp. 101–110). https://doi.org/10.1007/978-3-642-10439-8_11
Mendeley helps you to discover research relevant for your work.