AWRT: A Recommender System for Improving Revisiting Web Pages

Abstract

Web revisit has become one of the dominant phenomena in the web. Nevertheless, standard revisit features in typical web browsers currently do not significantly support this phenomenon. Various efforts, therefore, have been carried out to provide web users with approaches for facilitating web revisit. However, these approaches burden users with management tasks such as flagging and marking. This paper presents a novel approach, called the Adaptive Web Revisit Tool (AWRT), which helps facilitate revisiting web pages without involving users with any actions. In a similar way to the recommendation systems of well-known e-commerce websites, pages that are most likely to be revisited are adaptively recommended to users based on three attributes: frequency of visits, relation to current web page, and time per visit. AWRT consists of three main features: most visited pages (MVP), related pages (RP) and browsing behavior pages (BBP). AWRT was evaluated in terms of efficiency and usability using logging files and a satisfaction questionnaire. The results demonstrated an overall high rate of revisiting using AWRT when compared to a typical web browser i.e. Internet Explorer (IE) and a high users' satisfaction was demonstrated as well. Therefore, incorporating recommendation systems and adaptive techniques such AWRT can significantly improve web revisit.

Construction Mechanism of Digital Lifelong Learning Platform

“…To be specific, when a learner first enters into LDPSP, he is required to fill in personal career information, hobbies and professional information if he wants to become registered member of the platform. The platform collects information above to recommend corresponding learning contents for him [7,8]. If the learner has his own learning circle, the platform can recommend learning resource of his learning friends to him, which helps the learner quickly locate resources he needs.…”