Abstract
General method which allowed us to understand the target user was based on the graphs or tables using the distribution or average value of the research data. However, each person in charge of the processes in the development had often different imagination of target user, because of the difficulty of understanding the graphs or tables. So, persona method is focused on a method to present the imagination of the target users specifically. The virtual user or virtual users are created out based on the research data of target users. However, in the traditional persona method, there was a problem that the success of creating a persona depends on the skill of persona creator. In this study, we proposed a method for creating personas using the Bayesian Network Analysis, which allowed us to express the persona including not only the value of the research data but also the causal relationship between the data. Moreover, the proposed method is not inŽuenced by the skill or experience of persona creators and is most likely automated by adapting the persona template,REAS and Bayesian Network Analysis. We also validated the proposed method and evaluated the persona created by it.
