Personalized Video Search Based on User Log Mining

摘要

Since user log is an important carrier for recording the user?s behavior of Web search engine, by mining and analyzing user log the user?s laws and interest could be obtained effectively, which could be used to improve the accuracy of searching results when they use search engine. This study aimed to present an effective method for personalized video search based on user log mining. It first analyzed the related video log in user log which was free provided by Sogou company, classified them based on the contents, judged user?s intention by using sentence similarity matching algorithm and returned the videos which were closest to user?s search intention. The proposed method was verified through the experiment finally in this study.