An Online Recommender System for Large Web Sites
WI '04 Proceedings of the 2004 IEEE/WIC/ACM International Conference on Web Intelligence
An XML-based agent model for supporting user activities on the Web
Web Intelligence and Agent Systems
USE: A Concept-Based Recommendation System to Support Creative Search
WI-IAT '09 Proceedings of the 2009 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology - Volume 03
Hi-index | 0.00 |
The knowledge extracted from the analysis of historicalinformation of a web server can be used to develop personalizationor recommendation systems. Web Usage Mining(WUM) systems are specifically designed to carry out thistask by analyzing the data representing usage data about aparticular Web Site. Typically these systems are composedby two parts. One, executed offline, that analyze the serveraccess logs in order to find a suitable categorization, andanother executed online which is aimed at classifying theactive requests, according to the previous offline analysis.In this paper we propose a WUM recommendation system,implemented as a module of the Apache web server,that is able to dynamically generate suggestions to pagesthat have not yet been visited by a user and might be ofhis potential interest. Differently from previously proposedWUM systems, SUGGEST 2.0 incrementally builds andmaintain the historical information, without the need for anoffline component, by means of a novel incremental graphpartitioning algorithm. In the last part, we also analyze thequality of the suggestions generated and the performance ofthe module implemented. To this purpose we introduce alsoa new quality metric which try to estimate the effectivenessof a recommendation system as the capacity of anticipatingusers' requests that will be made farther in the future1.