PURE: Positive-Unlabeled Recommendation with Generative Adversarial Network

Recommender systems are powerful tools for information filtering with the ever-growing amount of online data. Despite its success and wide adoption in various web applications and personalized products, many existing recommender systems still suffer from multiple drawbacks such as large amount of unobserved feedback, poor model convergence, etc. These drawbacks of existing work are mainly due to the following two reasons: first, the widely used negative sampling strategy, which treats the unlabeled entries as negative samples, is invalid in real-world settings; second, all training samples are retrieved from the discrete observations, and the underlying true distribution of the users and items is not learned.

In this paper, we address these issues by developing a novel framework named PURE, which trains an unbiased positive-unlabeled discriminator to distinguish the true relevant user-item pairs against the ones that are non-relevant, and a generator that learns the underlying user-item continuous distribution. For a comprehensive comparison, we considered 14 popular baselines from 5 different categories of recommendation approaches. Extensive experiments on two public real-world data sets demonstrate that PURE achieves the best performance in terms of 8 ranking based evaluation metrics.