GAN-based Recommendation with Positive-Unlabeled Sampling

TL;DR

This paper introduces a GAN-based recommendation framework that employs positive-unlabeled sampling to improve recommendation accuracy, providing theoretical analysis and demonstrating superior performance on multiple datasets.

Contribution

It presents a novel GAN-based recommendation model integrating positive-unlabeled sampling, with theoretical bounds and empirical validation showing its effectiveness.

Findings

Outperforms 13 baseline methods on three datasets

Achieves higher ranking metrics across multiple evaluation criteria

Provides theoretical analysis of sampling bounds and convergence

Abstract

Recommender systems are popular tools for information retrieval tasks on a large variety of web applications and personalized products. In this work, we propose a Generative Adversarial Network based recommendation framework using a positive-unlabeled sampling strategy. Specifically, we utilize the generator to learn the continuous distribution of user-item tuples and design the discriminator to be a binary classifier that outputs the relevance score between each user and each item. Meanwhile, positive-unlabeled sampling is applied in the learning procedure of the discriminator. Theoretical bounds regarding positive-unlabeled sampling and optimalities of convergence for the discriminators and the generators are provided. We show the effectiveness and efficiency of our framework on three publicly accessible data sets with eight ranking-based evaluation metrics in comparison with thirteen popular baselines.