Modelling User Behaviour for Web Recommendation Using LDA Model

Web users exhibit a variety of navigational interests through clicking a sequence of Web pages. Analysis of Web usage data will lead to discover Web user access pattern and facilitate users locate more preferable Web pages via collaborative recommending technique. Meanwhile, latent semantic analysis techniques provide a powerful means to capture user access pattern and associated task space. In this paper, we propose a collaborative Web recommendation framework, which employs Latent Dirichlet Allocation (LDA) to model underlying topic-simplex space and discover the associations between user sessions and multiple topics via probability inference. Experiments conducted on real Website usage dataset show that this approach can achieve better recommendation accuracy in comparison to existing techniques. The discovered topic-simplex expression can also provide a better interpretation of user navigational preference.