Research on Clothing Recommendation Based on Clothing Transaction Data

Abstract

Clothing matching has great value. In order to solve the problem that the matching effect of clothing purchased by users is not very good, a method is proposed in this paper. When a user purchases a certain clothing, the matching clothing can be automatically recommended. In this study, the clothing is clustered according to the description text of the clothing features first, and then the association rules are mined for the clothing in the same cluster. Finally, by calculating the learning error, some ‘false matches’ are filtered out to improve the precision of the recommended clothing. According to the characteristics of clothing purchase data that the average purchase volume of clothing is small, the calculation process of Apriori algorithm is optimized. The experiments show that the proposed method can reduce the calculation time of association rules and increase the accuracy of clothing recommendation compared with the traditional association rule calculation.