User Personalized Recommendation Model Based on Web Log Mining
Yao Xuefeng · 2017 International Conference on Smart Grid and Electrical Automation (ICSGEA) · 2017
This article puts forward access frequency and browsing time-based browsing preference path mining algorithm as well as personalized information recommendation algorithm. At first, it necessarily performs pretreatment on original log file. Then, according to current users browsing behavior and the given mode in access mode base, that is, the mode in path mode base with higher browsing rate for match, the paper searches weight value with successful match in mixing matrix and provides users with recommending page service according to weight value arrangement in descending order. The decision-tree technology of data mining is applied to combine with users information search behavior to construct automatic recognition model of users information request. The log data mining and recommending test based on one commercial website show that this method effectively improves accuracy and coverage of personal resource recommendation.