Adaptive Online Retail Web Site Based on Hidden Markov Model
Author(s): WANG Shi, GAO Wen, HUANG Tie-Jun, MA Ji-yong, LI Jin-tao
Pages: 599-606
Year: 2001 Issue: 4
Journal: JOURNAL OF SOFTWARE
Keyword: Web数据挖掘; 隐马尔可夫模型; 关联规则; 自适应;
Abstract: There is a problem in online retail: the conflict between the different interests of all customers to different commodities and the commodity classification structure of Web site. This problem will make most customers access overabundant Web pages. To solve the problem, the Web page data, server data, and marketing data are mined to build a hidden Markov model. The authors use association rule discovery to get the large item set. Viterbi algorithm is used to find some paths that come from the root Web page to the Web page that the center of the large item set is in. This large item set is marked in the nodes that are in the paths. Through these steps, one can calculate all item sets and mark them in these paths. The large item sets will compete in the nodes for the limited space. Through this method the Web site will adjust itself to reduce the total access time of all users. This method can also be used in analysis of paths, advertisements, and reconstructing the Web site.
Pages: 599-606
Year: 2001 Issue: 4
Journal: JOURNAL OF SOFTWARE
Keyword: Web数据挖掘; 隐马尔可夫模型; 关联规则; 自适应;
Abstract: There is a problem in online retail: the conflict between the different interests of all customers to different commodities and the commodity classification structure of Web site. This problem will make most customers access overabundant Web pages. To solve the problem, the Web page data, server data, and marketing data are mined to build a hidden Markov model. The authors use association rule discovery to get the large item set. Viterbi algorithm is used to find some paths that come from the root Web page to the Web page that the center of the large item set is in. This large item set is marked in the nodes that are in the paths. Through these steps, one can calculate all item sets and mark them in these paths. The large item sets will compete in the nodes for the limited space. Through this method the Web site will adjust itself to reduce the total access time of all users. This method can also be used in analysis of paths, advertisements, and reconstructing the Web site.
Related Articles