Employing social network construction and analysis in web structure optimization

Abstract

The world wide web is growing continuously and rapidly; it is quickly facilitating the migration of tasks of the daily life into web-based. This trend shows time will come when everyone is forced to use the web for daily activities. Naive users arc the major concern of such a shift; so, it is necessary to have the web ready to serve them. We argue that this requires well optimized websites for users to quickly locate the information they arc looking for. This, on the other hand, becomes more and more important due to the widespread reliance on the many services available on the Internet nowadays. It is true that search engines can facilitate the task of finding the information one is looking for. However, search engines will never replace but do complement the optimization of a website's internal structure based on previously recorded user behavior. In this chapter, wc will present a novel approach for identifying problematic structures in websites. This method consists of two phases. The first phase compares user behavior, derived via web log mining techniques, to a combined analysis of the website's link structure obtained by applying three methods leading to more robust framework and hence strong and consistent outcome: (1) constructing and analyzing a social network of the pages constituting the website by considering both the structure and the usage information; (2) applying the Weighted PageRank algorithm; and (3) applying the Hypertext Induced Topic Selection (HITS) method. In the second phase, we use the term frequency-inverse document frequency (TFIDF) measure to investigate further the correlation between the page that contains the link and the linked to pages in order to further support the findings of the first phase of our approach. We will then show how to use these intermediate results in order to point out problematic website structures to the website owner. © 2010 Springer-Verlag Wien.