Showing papers by "Magdalini Eirinaki published in 2007"
TL;DR: UPR, a PageRank-style algorithm which combines usage data and link analysis techniques for assigning probabilities to Web pages based on their importance in the Web site's navigational graph, is presented and it is proved that this approach results in more objective and representative predictions than the ones produced from the pure usage-based approaches.
Abstract: The continuous growth in the size and use of the World Wide Web imposes new methods of design and development of online information services. The need for predicting the users' needs in order to improve the usability and user retention of a Web site is more than evident and can be addressed by personalizing it. Recommendation algorithms aim at proposing “next” pages to users based on their current visit and past users' navigational patterns. In the vast majority of related algorithms, however, only the usage data is used to produce recommendations, disregarding the structural properties of the Web graph. Thus important—in terms of PageRank authority score—pages may be underrated. In this work, we present UPR, a PageRank-style algorithm which combines usage data and link analysis techniques for assigning probabilities to Web pages based on their importance in the Web site's navigational graph. We propose the application of a localized version of UPR (l-UPR) to personalized navigational subgraphs for online Web page ranking and recommendation. Moreover, we propose a hybrid probabilistic predictive model based on Markov models and link analysis for assigning prior probabilities in a hybrid probabilistic model. We prove, through experimentation, that this approach results in more objective and representative predictions than the ones produced from the pure usage-based approaches.
25 citations
09 Nov 2007
TL;DR: The key feature of the methodology is that the personalization is based on the mining of navigation patterns extracted from previous user visits, which represent the navigation behaviour and interests of different users or user groups.
Abstract: Topic directories are popular means of organizing information resources in the web. In this work, we introduce a methodology for personalizing topic directories. The key feature of our methodology is that the personalization is based on the mining of navigation patterns extracted from previous user visits. These patterns, expressed in the form of visited categories and retrieved resources, represent the navigation behaviour and interests of different users or user groups. Our work provides a set of mining tasks for user navigation patterns and a set of personalization tasks that customize the organization of the topic directory according to these patterns for certain user groups.
12 citations