Personalized Recommendation for Public Transportation Using User Context

Abstract

This paper presents a new application of public transit guide system named PATRASH (Personalized Autonomous TRAansit recommendation System considering user context and History). Although existing transit guide systems provide a lot of functions, their users can not always use all the functions under their control because of the complexity of the functions. Under this situation, we developed and supply PATRASH to reduce complicated user operations. PATRASH provides simple functions users can easily use. The concept of PATRASH is “First Show, Second Search.” PATRASH displays routes and time table information recommended to the users. If they need, they can search another route. The target users of PATRASH are transit riders who come to Ito campus of Kyushu university. Ito campus locates in a rural area and has four routes run by three bus transportation companies and connected to a railroad or a subway. Since the main route usually used by each transit rider seems to be almost determined, we believe personalized function of PATRASH becomes more and more important. To realize the concept of PATRASH and give simple usability to the users, we propose a new 3D adaptive interface and release PATRASH to the public. We also present a preliminal case study of PATRASH project.