A Intelligent Method of Modelling Web User Interest

Xiyuan WU,Qinghua ZHENG,Ping WANG

DOI: https://doi.org/10.3969/j.issn.2095-6649.2014.9.07

2014-01-01

Abstract:Network user modeling is one of the key issues for network personalized service. The majority of the existed user modeling methods construct user model only from the perspective of the analysis of web log data user generated, but ignore the cognitive psychology's effect on the user model construction. This study combines some theories and technologies of data mining, knowledge management and cognitive psychology, and proposes a intelligent method of modeling web user interest, which consid-ers not only network log analysis but also user's cognitive characteristics. This method uses the extended topic map to describe the semantics of the network resources, provides statistical analysis on the user behavior log for accessing topic map, and introduces human memory law to build and update user interest model. Finally, this method can draw user models in concept level, knowledge level and other granularity. The empirical researches in personalized network English learning show that the proposed method is ef-fective to promote the network personalized study.