Mining interest navigation patterns based on hybrid Markov model

Authors:
Yijun Yu;Huaizhong Lin;Chun Chen
Affiliations:
Computer Institute, Zhejiang University, China;Computer Institute, Zhejiang University, China;Computer Institute, Zhejiang University, China
Venue:
ICCOMP'06 Proceedings of the 10th WSEAS international conference on Computers
Year:
2006

Citing 6
Cited 0

Footprints: history-rich tools for information foraging

Proceedings of the SIGCHI conference on Human Factors in Computing Systems
Link prediction and path analysis using Markov chains

Proceedings of the 9th international World Wide Web conference on Computer networks : the international journal of computer and telecommunications netowrking
An Heuristic to Capture Longer User Web Navigation Patterns

EC-WEB '00 Proceedings of the First International Conference on Electronic Commerce and Web Technologies
Design, implementation, and evaluation of duplicate transfer detection in HTTP

NSDI'04 Proceedings of the 1st conference on Symposium on Networked Systems Design and Implementation - Volume 1
SNIF-ACT: a model of information foraging on the world wide web

UM'03 Proceedings of the 9th international conference on User modeling
Generating dynamic higher-order markov models in web usage mining

PKDD'05 Proceedings of the 9th European conference on Principles and Practice of Knowledge Discovery in Databases

Quantified Score

Hi-index	0.00

Visualization

Abstract

Each user accesses a Web site with certain interest. The interest is associated with his navigation patterns. The interest navigation patterns represent different interest of the users. In this paper, hybrid Markov model is proposed for interest navigation pattern discovery. The novel model is better in prediction overlay rate and prediction correct rate than traditional Markov models. User group interest is also defined in this paper. The probability of user group interest navigation from one page to another is computed by navigation path characteristics and time characteristics. Compared with the previous ones, the results of the experiment show that the performance is improved efficiently by the hybrid Markov model.