Abstract:Incorporating Knowledge Graphs (KG) into recommeder system has attracted considerable attention. Recently, the technical trend of Knowledge-aware Recommendation (KGR) is to develop end-to-end models based on graph neural networks (GNNs). However, the extremely sparse user-item interactions significantly degrade the performance of the GNN-based models, as: 1) the sparse interaction, means inadequate supervision signals and limits the supervised GNN-based models; 2) the combination of sparse interactions (CF part) and redundant KG facts (KG part) results in an unbalanced information utilization. Besides, the GNN paradigm aggregates local neighbors for node representation learning, while ignoring the non-local KG facts and making the knowledge extraction insufficient. Inspired by the recent success of contrastive learning in mining supervised signals from data itself, in this paper, we focus on exploring contrastive learning in KGR and propose a novel multi-level interactive contrastive learning mechanism. Different from traditional contrastive learning methods which contrast nodes of two generated graph views, interactive contrastive mechanism conducts layer-wise self-supervised learning by contrasting layers of different parts within graphs, which is also an "interaction" action. Specifically, we first construct local and non-local graphs for user/item in KG, exploring more KG facts for KGR. Then an intra-graph level interactive contrastive learning is performed within each graph, which contrasts layers of the CF and KG parts, for more consistent information leveraging. Besides, an inter-graph level interactive contrastive learning is performed between the local and non-local graphs, for sufficiently and coherently extracting non-local KG signals. Extensive experiments conducted on three benchmark datasets show the superior performance of our proposed method over the state-of-the-arts.

Multitype view of knowledge contrastive learning for recommendation

Multi-level Cross-view Contrastive Learning for Knowledge-aware Recommender System

Knowledge-Adaptive Contrastive Learning for Recommendation

Attribute Mining Multi-View Contrastive Learning Network for Recommendation

Multi-contrastive Learning Recommendation Combined with Knowledge Graph

Knowledge filter contrastive learning for recommendation

Multi-modal Recommendation Based on Knowledge Graph

Improving Knowledge-aware Recommendation with Multi-level Interactive Contrastive Learning.

Hierarchically Coupled View-Crossing Contrastive Learning for Knowledge Enhanced Recommendation

Enhanced knowledge graph recommendation algorithm based on multi-level contrastive learning

Multi-view Contrastive Learning for Knowledge-Aware Recommendation

Knowledge-aware Multi-view Cross Learning for Edge-based Collaborative Recommendation

Knowledge-based Recommendation with Contrastive Learning

Knowledge-enhanced Multi-View Graph Neural Networks for Session-based Recommendation

Cross-modal Knowledge Graph Contrastive Learning for Machine Learning Method Recommendation

Enhancing Recommendations with Contrastive Learning from Collaborative Knowledge Graph.

Exploring Multi-dimension User-Item Interactions with Attentional Knowledge Graph Neural Networks for Recommendation

Mixed-curvature Knowledge-Enhanced Graph Contrastive Learning for Recommendation.

MI-KGNN - Exploring Multi-dimension Interactions for Recommendation Based on Knowledge Graph Neural Networks.

Disentangled Contrastive Learning for Knowledge-Aware Recommender System

Enhancing Knowledge-Aware Recommendation with Contrastive Learning.