Integrated Web Recommendation Model with Improved Weighted Association Rule Mining

Web log data is essential for improving the performance of the Web. It contains large, heterogeneous and diverse data. However, analyzing Web log data is a tedious process for Web developers, Web designers, technologists and end users. In this work, a new weighted association mining algorithm is developed to identify the best association rules that are useful for Web site restructuring and to give recommendations that reduce false visits and improve users' navigation behavior. The algorithm finds the frequent item set from a large uncertain database. The frequent scanning of the database in each time is the problem with the existing algorithms which leads to a complex output set and time consuming process. The proposed algorithm scans the database only once at the beginning of the process and the generated frequent item sets, which are stored into the database. Evaluation parameters such as support, confidence, lift and number of rules are considered to analyze the performance of the proposed algorithm and traditional association mining algorithm.