Interactive Personalized Catalogue for M-Commerce

Abstract

M-commerce possesses two distinctive characteristics that distinguish it from traditional e-commerce: the mobile setting and the small form factor of mobile devices. Of these, the size of a mobile device will remain largely unchanged due to the tradeoff between size and portability. Small screen size and limited input capabilities pose a great challenge for developers to conceptualize user interfaces that have good usability while working within the size constraints of the device. In response to the limited screen size of mobile devices, there has been unspoken consensus that certain tools must be made available to aid users in coping with the relatively large volume of information. Recommender systems have been proposed to narrow down choices before presenting them to the user (Feldman, 2000). We propose a product catalogue where browsing is directed by an integrated recommender system. The recommender system is to take incremental feedback in return for browsing assistance. Product appearance in the catalogue will be dynamically determined at runtime based on user preference detected by the recommender system. The design of our hybrid m-commerce catalogue recommender system investigated the typical constraints of m-commerce applications to conceptualize a suitable catalogue interface. The scope was restricted to the case of having personal digital assistant (PDA) as the mobile device. Thereafter, a preference detection technique was developed to serve as the recommender layer of the system.