Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Graph Cross-Correlated Network for Recommendation

Hao Chen, Yuanchen Bei, Wenbing Huang, Shengyuan Chen, Feiran Huang, Xiao Huang

TL;DR

The Graph Cross-correlated Network for Recommendation (GCR) is proposed, which serves as a general recommendation paradigm that explicitly considers correlations between user/item subgraphs and comprehensively incorporates the cross-correlated terms for recommendations.

Abstract

Collaborative filtering (CF) models have demonstrated remarkable performance in recommender systems, which represent users and items as embedding vectors. Recently, due to the powerful modeling capability of graph neural networks for user-item interaction graphs, graph-based CF models have gained increasing attention. They encode each user/item and its subgraph into a single super vector by combining graph embeddings after each graph convolution. However, each hop of the neighbor in the user-item subgraphs carries a specific semantic meaning. Encoding all subgraph information into single vectors and inferring user-item relations with dot products can weaken the semantic information between user and item subgraphs, thus leaving untapped potential. Exploiting this untapped potential provides insight into improving performance for existing recommendation models. To this end, we propose the Graph Cross-correlated Network for Recommendation (GCR), which serves as a general recommendation paradigm that explicitly considers correlations between user/item subgraphs. GCR first introduces the Plain Graph Representation (PGR) to extract information directly from each hop of neighbors into corresponding PGR vectors. Then, GCR develops Cross-Correlated Aggregation (CCA) to construct possible cross-correlated terms between PGR vectors of user/item subgraphs. Finally, GCR comprehensively incorporates the cross-correlated terms for recommendations. Experimental results show that GCR outperforms state-of-the-art models on both interaction prediction and click-through rate prediction tasks.

Graph Cross-Correlated Network for Recommendation

TL;DR

The Graph Cross-correlated Network for Recommendation (GCR) is proposed, which serves as a general recommendation paradigm that explicitly considers correlations between user/item subgraphs and comprehensively incorporates the cross-correlated terms for recommendations.

Abstract

Collaborative filtering (CF) models have demonstrated remarkable performance in recommender systems, which represent users and items as embedding vectors. Recently, due to the powerful modeling capability of graph neural networks for user-item interaction graphs, graph-based CF models have gained increasing attention. They encode each user/item and its subgraph into a single super vector by combining graph embeddings after each graph convolution. However, each hop of the neighbor in the user-item subgraphs carries a specific semantic meaning. Encoding all subgraph information into single vectors and inferring user-item relations with dot products can weaken the semantic information between user and item subgraphs, thus leaving untapped potential. Exploiting this untapped potential provides insight into improving performance for existing recommendation models. To this end, we propose the Graph Cross-correlated Network for Recommendation (GCR), which serves as a general recommendation paradigm that explicitly considers correlations between user/item subgraphs. GCR first introduces the Plain Graph Representation (PGR) to extract information directly from each hop of neighbors into corresponding PGR vectors. Then, GCR develops Cross-Correlated Aggregation (CCA) to construct possible cross-correlated terms between PGR vectors of user/item subgraphs. Finally, GCR comprehensively incorporates the cross-correlated terms for recommendations. Experimental results show that GCR outperforms state-of-the-art models on both interaction prediction and click-through rate prediction tasks.

Paper Structure

This paper contains 22 sections, 15 equations, 4 figures, 6 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. RELATED WORK
  3. Collaborative Filtering
  4. Graph Neural Networks for Recommendations
  5. Click-Through Rate Prediction
  6. Preliminaries
  7. Methodology
  8. Plain Graph Representation (PGR)
  9. Cross-Correlated Aggregation (CCA)
  10. Optimization
  11. experiments
  12. Experimental Setup
  13. Datasets
  14. Comparing Methods
  15. Evaluation Details
  16. ...and 7 more sections

Figures (4)

  • Figure 1: An illustration of the user-item graph and the different semantic meanings of each hop of neighbors in the user/item subgraph.
  • Figure 2: (a) & (b): Sketch of existing recommendation models; (c): The overall framework architecture of our proposed GCR. (a) CF-based recommendation models utilize the dot product of the learned user/item embeddings to infer user-item interactions. (b) GNN-based recommendation models begin by performing graph convolutions to generate embeddings for multi-hop neighbors. These embeddings are then combined to construct super vectors, which are used for making predictions. (c) GCR first extracts graph embeddings with Plain Graph Representation and then flexibly considers all potential cross-correlations between the target user and the target item with Cross-Correlated Aggregation to infer their interaction.
  • Figure 3: Comparison of inferring times on three experimental datasets.
  • Figure 4: Visualization of the weights of the cross-correlation terms. For HCC, we have single-weight parameters for these terms (red lines), while we have 64 weight parameters for these terms of ECC (blue lines).

Theorems & Definitions (1)

  • Definition 1: Recommendation Task