摘要

This paper presents a novel and effective approach for constructing user profile based on Extended Probabilistic Latent Semantic Analysis (E-PLSA). In the approach, firstly, an initial user profile is created through analyzing web usage information based on E-PLSA, and then the user profile is learned and updated through incrementally analyzing user's queries. The experimental results demonstrate that the performance of E-PLSA method outperformed Probabilistic Latent Semantic Analysis (PLSA) method in satisfy users' personalized information requirements.

  • 出版日期2007-8
  • 单位沈阳大学