Automatic classification for grouping designs in fashion design recommendation agent system

Abstract

Nowadays, users spend much time and effort in finding the best suitable designs since more and more information is placed on-line. To save their time and effort in searching the designs they want, the user-adapting recommendation system is required. In this paper, Automatic Classification for Grouping designs (ACG) in a Fashion Design Recommendation Agent System (FDRAS) is proposed. The ACG algorithm groups designs into clusters based on these classified designs. It is possible that if the design requires simultaneous regrouping in all other groups, the ACG algorithm can be used to improve efficiency of information retrieval and sorting, in the FDRAS datasets. The proposed method is evaluated on a large database, significantly outperforming the nearest-neighbor model and k-mean clustering in the prototype user-adapting FDRAS. This method can solve the large-scale dataset problem without deteriorating accuracy quality. © Springer-Verlag Berlin Heidelberg 2006.