Mining Users' Similarity from Moving Trajectories for Mobile E-commerce Recommendation
Haidong Zhong, Shaozhong Zhang, Yanling Wang, Shifeng Weng, Yonggang Shu · International Journal of Hybrid Information Technology · 2014
Users’ similarity mining in mobile e-commerce systems is an important field with wide applications, such as personalized recommendation and accurate advertising. Moving trajectories of e-commerce users contain much useful information, providing a very good opportunity for understanding the users’ interesting and discovering the similarity between mobile-device-holders. In this paper, we explores the problems in the existing mobile ecommerce recommendation methods, and propose a mobile users’ moving trajectories mining based user similarity discovering approach for mobile e-commerce system. We formally defines the moving trajectory and view the areas, where users stay within for a certain time, as interested regions, which reflect the preferences of mobile-device-holders. Based on the number of overlapped interested areas, a user similarity measure method is proposed. Experimental evaluation, conducted based on the publicly available datasets commendably demonstrate the effectiveness of our approach.