신한복 커스터마이징 온라인 플랫폼에 적합한 의상 추천 알고리즘 연구

Abstract

Recently, as idol singers and other celebrities wearing various types of new hanbok have been shown on TV, public interest in the new hanbok is increasing. Also, with the continuous spread of the Korean Wave, global interest in the new Hanbok is also increasing. Today, the proportion of online clothing industry is increasing due to the development of the IT industry and the improvement of the performance of mobile devices. According to data released by the National Statistical Office in 2020, the size of the domestic online fashion market grew by 13.4% compared to the previous year to reach 42,687.9 billion won. This phenomenon has accelerated after the corona pandemic, and there is a forecast that in 2021, the transaction amount of the major online fashion platform Big 5 will grow by more than 40% compared to 2020. On the other hand, the volume of transactions in offline stores, including department stores, continues to decrease. In online shopping, a recommendation algorithm is a basic method to increase a user's purchasing power. A representative recommendation algorithm is collaborative filtering. However, there is a disadvantage that a product rating is absolutely necessary to apply collaborative filtering. In particular, the accumulated data for new Hanbok is insufficient because the market has not been activated yet. Therefore, this paper proposes a recommendation method suitable for developing a new hanbok online customizing platform based on the analysis results of existing studies. The method proposed in this paper can be used as an initial recommendation algorithm for the new hanbok online customizing platform.