Work

ENSG: Enhancing Negative Sampling in Graph Convolutional Networks for Recommendation Systems

Year: 2024

Type: article

Abstract: In the field of recommender systems, difficult negative samples that are easily misclassified by classifiers have higher learning value. In order to solve the problem that the existing graph convolutional network lacks the exploration of negative sampling strategy, a multivariate sampling based on graph convolution collaborative filtering recommendation model was proposed. Firstly, part of the information in the positive sample was injected into the negative sample, and the high quality difficult negative sample was selected by the inner product method. Then, contrastive learning is used to maximize the similarity between the difficult negative samples and the positive samples, so that the difficult negative samples are as close to the positive samples as possible in the feature space. Ultimately, a multi-task approach was employed to concurrently enhance both the recommendation supervision task and the contrastive learning task. To validate the effectiveness of our approach, we conducted comparative evaluations utilizing three publicly accessible datasets. The experimental outcomes indicate that our proposed model surpasses the baseline model in terms of performance.

Source: Electronics

DOI: 10.3390/electronics13234696

Authors: Hai Yan, Zheng Ji Tao, Liu Zhi Zhong, Wang Dong Yang, Ding Cheng Rui

Institutions: North China University of Water Resources and Electric Power, Yantai University

Language: English

Topic: Recommender Systems and Techniques

Subfield: Information Systems

Field: Computer Science

Domain: Physical Sciences

Open Access status: gold

Estimated APC paid: $2,728

Funder: National Natural Science Foundation of China