Neural Pairwise Ranking Factorization Machine for Item Recommendation

The factorization machine models attract significant attention from academia and industry because they can model the context information and improve the performance of recommendation. However, traditional factorization machine models generally adopt the point-wise learning method to learn the model parameters as well as only model the linear interactions between features. They fail to capture the complex interactions among features, which degrades the performance of factorization machine models. In this paper, we propose a neural pairwise ranking factorization machine for item recommendation, which integrates the multi-layer perceptual neural networks into the pairwise ranking factorization machine model. Specifically, to capture the high-order and nonlinear interactions among features, we stack a multi-layer perceptual neural network over the bi-interaction layer, which encodes the second-order interactions between features. Moreover, the pair-wise ranking model is adopted to learn the relative preferences of users rather than predict the absolute scores. Experimental results on real world datasets show that our proposed neural pairwise ranking factorization machine outperforms the traditional factorization machine models.