Mining Users Interest Navigation Patterns Using Improved Ant Colony Optimization
Xuyang Wei, Yan Wang, Zhongliang Li, Tengfei Zou, Guocai Yang · Intelligent Automation & Soft Computing · 2015
Web log mining is mainly to acquire users’ interest navigation patterns from web logs and has been the subject of the web personalization research. In this paper, we define a new concept “interest pheromone” and present a group users’ navigation paths model. Then we propose a simple algorithm based on improved Ant Colony Optimization (ACO) to mine users’ dynamic interest. In this algorithm, three factors relative browsing time, access frequency and operation time are considered to measure the “interest pheromone”, which better reflects users’ real interest. Finally, we conduct the simulation experiments to contrast the accuracy of navigation patterns mined by our approach and existing approaches. Experimental results illustrate that the proposed paradigm can truly capture users’ browsing preference effectively.