A Profiling Based on Categorical Inner Products in Quantification Theory Type 3

抄録

In order to estimate the profiles of users, the relationship between the groups such as ages and/or gender and the attributes such as hobbies and/or preference which are acquired by the response of a questionnaire are analyzed. The aim of this paper is to obtain the detail relationship between each category. This paper focused on the independency of the axes in the score space in quantification theory type 3. In this paper, the information of the groups and the attributes acquired from a questionnaire data were converted from the feature space into the score space by quantification theory type 3. The profiling method based on the categorical inner products between groups and attributes and the visualization method for the analysis were proposed and were applied to an actual questionnaire data in this paper.