Clustering-Based User Knowledge Extraction Structure for Furniture Design

Furniture, a part of our daily lives, is designed in various forms and textures according to users' needs. Differences in furniture designs directly impact users' feelings and, therefore, their experiences with the product. The card sorting technique is one of the user experience research methods. This method is used to create mind maps of users towards a product. This study investigates user knowledge structure for furniture designs using card sorting and hierarchical cluster analysis methods. Card sorting data reveals patterns in how user experience designers categorize the semantic words users select for furniture designs through hierarchical cluster analysis. Hierarchical clustering, a popular data analysis method, has been used to detect semantic or contextual similarities and relationships of semantic words. Dendrogram diagrams were created using the hierarchical clustering method. With dendrogram diagrams, the relationship between the semantic word data related to furniture, how they are clustered, and how the clusters are grouped at lower and upper levels are examined. Thus, it helps designers to develop more optimal, user-centered product design. In this study, after the semantic space of 102 semantic words used in furniture comfort design was created, it was reduced to 35 words by the brainstorming method. Then, the semantic words were clustered using the hierarchical clustering algorithm by applying the card sorting technique to 69 participants. According to the results obtained, the rankings for furniture designs are clustered under five clusters: beautiful - ugly, multifunctional - single function, elegant - coarse, living - nonliving, and rough - smooth. The study provided data that designers can use to redesign furniture in a user-centered way.