Abstract: Intelligent discovery of users' navigation pattern has been a hot research issue in the E-Commerce field in recent years. In this paper, users' access interests are defined by combining the information of users' time duration on a page with the keywords on the pages in the E-Commerce Website. The Interest Navigational Path Model is constructed based on the classical Hidden Markov Chains Model. Next, the discovery method for user's interest navigational paths and corresponding mining algorithm are presented. Finally, the experiments are conducted with simulative data, real datasets collected from an online Bookselling B-to-C E-commerce site. Furthermore, the comparative experiment with a classical algorithm is conducted. The experimental results show that the presented model and algorithm can accurately and efficiently find the paths information associated with users' access interests. The method can be adopted as a more effective and practical tool for intelligent navigation discovery oriented to the E-Commerce field.

Key words: intelligent E-Commerce, HMM, web data mining, Interest Navigation Patterns