关键词: 网购, 尺码推荐, 松量阈值, 服装廓形, 合体性

Abstract: In view that consumers cannot try on clothing purchased online, resulting in poor fitness and high refund of clothing, with women's T-shirts as an example, through literature collection, expert interviews and empirical research, the clothing fitness ease threshold of the key parts of clothing and the judging criteria for different silhouettes were determined, and a complete size recommendation system was established on this basis, which stores information such as clothing specifications and dimensions, user' personal information in the system database. The system front end was designed using bootstrap framework and the back end was developed using C# language. Adopting quadratic function as the core algorithm, the system is feasible to realize intelligent recommendation of clothing size by calculating and comparing the function values of different sizes. Through experimental verification, it is proved that the accuracy rate of this recommendation system can exceed 88%. It is conductive to eliminating consumers' shopping doubts to some extent, lowering logistics costs, and promoting the development of apparel e-commerce.

Key words: online shopping, size recommendation, clothing ease threshold, garment silhouette, fitness