2023 年 59 巻 Supplement 号 p. JS-01
This study aims to explore the classification of lower body types among Korean women using the Fuzzy C-means clustering approach. It is important to investigate various shapes of the human body in product design. Previously, most studies focused on specific body parts or targeting specific user groups, and considered variations based on factors such as gender, age and so on. We collected data which is related to the lower body of Korean females and removed missing values and outliers. Various components are reduced by utilizing factor analysis in preprocessing with KMO and Bartlett’s test of sphericity. After that, we performed the Fuzzy C- means clustering with 5 centers. Four factors were derived from factor analysis. After clustering, five types were found each characteristic of the lower body of Korean females as well. In addition, the characteristics of the lower body by age were investigated. This study identifies the probability distribution of belonging to a cluster according to anthropometric items. The comprehension of different lower body types of Korean women is helpful in developing product design. Furthermore, this method can be applied not only for the classification of the upper body type but also for the classification of the Korean human body type.
