User modeling for car selection based on probabilistic methods

Abstract

User modeling becomes more important at product making business because companies have to deal with a wide variety of customers favorites. We are researching customer behavior modeling based on customers' sensibility or "kansei". In this study, we focused on car selection as the first case of customer modeling and we researched factors related to the selection of a car through qualitative surveys. We took an Internet survey for car selection based on the result of past surveys. In this paper, we report the result obtained from an analysis of the survey data.