Research on the construction method of knowledge graph for clothing recommendation based on expert knowledge

Honglei Zhou,Xun Liu,Yuxuan Fang,Bo Zhang
DOI: https://doi.org/10.1109/cisai54367.2021.00217
2021-09-01
Abstract:In response to the lack of professional guidance in the field of clothing recommendation, a method of constructing a knowledge graph in the field of clothing recommendation based on expert knowledge is proposed. By summarizing the concepts of the apparel recommendation domain and drawing on the knowledge of image design experts, a system of apparel recommendation rules was constructed. The knowledge graph was further constructed and expanded under the guidance of the clothing rules system, and the application of the joint information extraction model based on DGCNN and probabilistic graph in the field of clothing recommendation was studied, and the storage form of knowledge graph is introduced. Experiments show that the model has a high accuracy rate in the scenarios of this paper. Combined with the construction process and experimental results of this paper, it is shown that the construction method proposed in this paper has some practical value for application in the field of clothing recommendation.
What problem does this paper attempt to address?