Extending web content management systems navigation capabilities with semantic navigation maps

Abstract

This paper presents an automatic approach built on information retrieval and clustering techniques to enhance the navigation capabilities of modern Web Content Management Systems (WCMSs). The approach uses Latent Semantic Indexing to discover correlations between the contents published through these systems, and a fuzzy clustering algorithm to form groups of related contents. For each page of the developed website, a set of navigation links towards pages showing similar or related content and a measure of such similarity is proposed. An implementation of the approach for the Joomla! Open Source WCMS and the results from a case study on a real world website are also presented.