Data Preprocessing Method Based on User Characteristic of Interests for Web Log Mining

Abstract

Web log mining is the most important method in Web data mining, and data preprocessing is the primary work. In order to find more value access mode and reduce the data size from the Web, find the data of users and even between users, this paper puts forward a method of Web log data preprocessing based on user characteristic of interests, and then put forward some concepts such as user interest, user interest similarity. Finally, after some experiments, we can show the superiority and recommended value of this new method.