Display online store products by automatic merchandise classification based on product similarity

This study presents an automatic merchandise classification approach based on similarity of product attributes. The proposed approach can automatically separate products into appropriate groups, allow consumers to rapidly and conveniently find products with similar attributes, and considerably reduce the workload associated with managing Internet stores. To illustrate the usefulness of the automatic merchandise classification method proposed here, this study includes a prototype website for making a practical comparison between this method and existing classification methods. An experiment involving 205 participants revealed that this method not only increased the efficiency of website management, but also permitted consumers to easily identify the products they required. Participants also indicated greater satisfaction when using the method proposed in this study.