Geometric Matrix Completion with Recurrent Multi-Graph Neural Networks
This work addresses recommender systems by improving matrix completion with graph-based methods, offering a more efficient and effective solution, though it appears incremental as it builds on existing graph and neural network techniques.
The paper tackles the problem of matrix completion for recommender systems by introducing a geometric deep learning approach that combines graph convolutional and recurrent neural networks to exploit local stationarity structures and reduce parameter count. The method outperforms state-of-the-art techniques on synthetic and real datasets.
Matrix completion models are among the most common formulations of recommender systems. Recent works have showed a boost of performance of these techniques when introducing the pairwise relationships between users/items in the form of graphs, and imposing smoothness priors on these graphs. However, such techniques do not fully exploit the local stationarity structures of user/item graphs, and the number of parameters to learn is linear w.r.t. the number of users and items. We propose a novel approach to overcome these limitations by using geometric deep learning on graphs. Our matrix completion architecture combines graph convolutional neural networks and recurrent neural networks to learn meaningful statistical graph-structured patterns and the non-linear diffusion process that generates the known ratings. This neural network system requires a constant number of parameters independent of the matrix size. We apply our method on both synthetic and real datasets, showing that it outperforms state-of-the-art techniques.