Deep Collaborative Filtering with Multi-Aspect Information in Heterogeneous Networks

Abstract

Recently, recommender systems play a pivotal role in alleviating the problem of information overload. Latent factor models have been widely used for recommendation. Most existing latent factor models mainly utilize the interaction information between users and items, although some recently extended models utilize some auxiliary information to learn a unified latent factor for users and items. The unified latent factor only represents the characteristics of users and the properties of items from the aspect of purchase history. However, the characteristics of users and the properties of items may stem from different aspects, e.g., the brand-aspect and category-aspect of items. Moreover, the latent factor models usually use the shallow projection, which cannot capture the characteristics of users and items well. Deep neural network has shown tremendous potential to model the non-linearity relationship between users and items. It can be used to replace shallow projection to model the complex correlation between users and items. In this paper, we propose a Neural network based Aspect-level Collaborative Filtering model (NeuACF) to exploit different aspect latent factors. Through modelling the rich object properties and relations in recommender system as a heterogeneous information network, NeuACF first extracts different aspect-level similarity matrices of users and items respectively through different meta-paths, and then feeds an elaborately designed deep neural network with these matrices to learn aspect-level latent factors. Finally, the aspect-level latent factors are fused for the top-N recommendation. Moreover, to fuse information from different aspects more effectively, we further propose NeuACF++ to fuse aspect-level latent factors with self-attention mechanism. Extensive experiments on three real world datasets show that NeuACF and NeuACF++ significantly outperform both existing latent factor models and recent neural network models.

Light Graph Convolutional Collaborative Filtering With Multi-Aspect Information

“…DMF [21] takes the interaction history of the user and item as a feature vector and inputs it to a multi-layer perceptron to learn the latent expression of user and item. NeuACF [22] introduces aspect-level information on the framework of NCF and designs an attention mechanism for learning the weight of aspect-level information. It should be noted that our method is different from NeuACF, which adopts meta-path to extract similarity matrix of users and items from heterogeneous graph.…”

Heterogeneous Neural Collaborative Filtering for a Business Recommender System

“…The NeuCF model was implemented [18] to test this dataset. Both HR@15 and NDCG@15 show that HNCF outperformed NeuCF.…”

Heterogeneous Information Network-based Interest Composition with Graph Neural Network for Recommendation

“…In the former step, diverse information has been mined with various technologies. For example, the matrix factorization is utilized in HeteRec [29], FMG [?, fmg]nd HueRec [26] to accomplish this step, while the HNAFM [1] and NeuACF [17] adopt MLP to extract deeper information of users and items according to their initialized features in each metapath. Besides, the HERec [18] and HopRec [6] capture the graph structure in corresponding meta-path with the help of DeepWalk [13].…”

“…Generally, most of HIN-based RSs [29] [33], [17] contain two stage, i.e., extracting information from multi meta-paths and fusing them to recommendation. In the former stage, the RS mines useful information in each meta-path through different methods such as matrix factorization and multi-layer perceptron (MLP).…”