Toward Personalized Public Transportation Recommendation System with Adaptive User Interface

Public transportation guidance services, such as Yahoo, Jorudan and NAVITIME, are widely used nowadays and support our daily lives. Although they provide useful services, they have not fully been personalized yet. This paper presents a personalized transportation system called PATRASH: Personalized Autonomous TRAnsportation recommendation System considering user context and History. In particular, we discuss an Adaptive User Interface (AUI) of Patrash. Before designing a personalized route recommendation function for Patrash´s AUI, we investigated possibilities and effectiveness of the function. First, we collected and analyzed 10 subjects´ usage histories of public transportation. Through this investigation, we confirmed the possibilities and effectiveness of the personalized route recommendation function. Second, we investigated the effectiveness of the basic functions of Patrash´s AUI by comparing with two major transportation guidance systems in Japan. We evaluated those systems from the points of views of usabilities: click costs and time costs. The experimental results illustrate the effectiveness of AUI of Patrash.