A HowNet Based Web Log Mining Algorithm

摘要

Web log mining is used to extract user access pattern. An algorithm is proposed in this paper to resolve the problem of bad explanation for page sequence of web log mining. The algorithm firstly transforms user visited page sequence into maximal forward sequence, and then uses HowNet based semantic similar algorithm to describe user interest in visit sequence and explains the interest movement with certain semantic words. The algorithm will help web sites provide personalization pages.