Differentiated Fashion Recommendation Using Knowledge Graph and Data Augmentation

Cairong Yan,Yizhou Chen,Lingjie Zhou

doi:10.1109/access.2019.2928848

Cairong Yan, Yizhou Chen + Show 1 more

Open Access

https://doi.org/10.1109/access.2019.2928848

Copy DOI

Abstract

E-commerce recommender systems (RSs) can help users quickly find what they need or new products they might be interested in. To continuously enhance user trust in the website, improve page visits and dwell time, and most importantly, increase gross merchandise value (GMV), it is crucial to understand and capture the important information hidden in the data, which has a great impact on user choice. The fashion e-commerce websites can collect the attributes of items and users as well as the user purchase behaviors, but lack the fine-grained classification of the items and the implicit relationship between items and users. This paper focuses on Amazon fashion dataset, one of the most widely used datasets in the fashion field. A differentiated recommendation framework is proposed that provides different recommendation paths for active and inactive users to improve the overall recommendation quality. In the framework, a data augmentation algorithm based on transfer learning is proposed to filter out the irrelevant items and label items with fine-grained tags, and a user-item knowledge graph is built to discover the potential relationship between items and users. Finally, a differentiated recommendation strategy is put forward to make different recommendations for users with different characteristics. The experimental results show that through data augmentation algorithm to improve data quality, factorization machine model produces higher recommendation accuracy, the constructed knowledge graph can alleviate the cold start problem for recommendation, and the differentiated recommendation strategy has achieved better recommendations for both active and inactive users.

Highlights

E-commerce recommender systems (RSs) are personalized recommendation tools to enhance the overall marketing performance of e-commerce platforms
We will try to improve the accuracy and alleviate the cold start problem, and explore a new framework to deal with both issues at the same time
We propose the idea of using knowledge graph to alleviate the cold start problem, and build a user-item knowledge graph on the fashion dataset to explore the implicit relationship between users and items

Summary

INTRODUCTION

E-commerce recommender systems (RSs) are personalized recommendation tools to enhance the overall marketing performance of e-commerce platforms. Ayala-Gomez et al proposed to use knowledge graph to extend the semantic features of a given abstract and combine them with other features such as indegree and recency to fit the learning ranking model, which is used to generate citation recommendation in academic search engines [27]. This paper attempts to build a knowledge graph in a real e-commerce dataset, in order to provide a way to alleviate the cold start problem of recommendation tasks. StyleNet is a style representation learning model based on the deep neural network, which is proposed in our previous work [30] It is a multi-tasking clothing classification framework that can divide items by season, type, and style.

RECOMMENDATION WITH FM

Findings

CONCLUSION