@article{414,
  author = {Tich Phuoc Tran, Thi Thanh Sang Nguyen, Kien Cuong Dang, Xiaoying Kong},
  title = {An Efficient Web-Page Recommender System using Frequent Pattern Discovery and Dynamic Markov Models},
  journal = {International Journal of Web Applications},
  year = {2011},
  volume = {3},
  number = {1},
  doi = {},
  url = {http://www.dline.info/ijwa/fulltext/v3n1/1.pdf},
  abstract = {The Internet has recently become not only one of the most popular communication channels but also the most accessible and searchable information repository of different domains. Billions of users surf the Internet everyday to search for information or visit social network and e-commerce websites. The web usage behaviors of these users can be analyzed by Web Usage Mining (WUM) systems to discover useful knowledge that helps improve service performance. Despite recent successes, existing WUM systems still cannot cope with the growing dynamics and complexity of the Web, resulting in overwhelming overheads and low efficiency. In this paper, an innovative Web-Page Recommender System is proposed to model user web browsing behaviors, extract popular web paths and predict web navigation possibilities. Particularly, the main inference algorithm in this system integrates advanced Frequent Pattern Recognition methods and Stochastic Markov Models to achieve an optimal balance between superior predictive accuracy and excessively demanding computation in higher order models. Empirical analysis suggests that our system outperforms other conventional methods with respect to complexity reduction and accuracy improvement.},
}