Towards Low-loss 1-bit Quantization of User-item Representations for Top-K Recommendation

Due to the promising advantages in space compression and inference acceleration, quantized representation learning for recommender systems has become an emerging research direction recently. As the target is to embed latent features in the discrete embedding space, developing quantization for user-item representations with a few low-precision integers confronts the challenge of high information loss, thus leading to unsatisfactory performance in Top-K recommendation. In this work, we study the problem of representation learning for recommendation with 1-bit quantization. We propose a model named Low-loss Quantized Graph Convolutional Network (L2Q-GCN). Different from previous work that plugs quantization as the final encoder of user-item embeddings, L2Q-GCN learns the quantized representations whilst capturing the structural information of user-item interaction graphs at different semantic levels. This achieves the substantial retention of intermediate interactive information, alleviating the feature smoothing issue for ranking caused by numerical quantization. To further improve the model performance, we also present an advanced solution named LQ-GCN0=; with quantization approximation and annealing training strategy. We conduct extensive experiments on four benchmarks over Top-K recommendation task. The experimental results show that, with nearly 9× representation storage compression, LQ-GCN0=; attains about 90∼99% performance recovery compared to the state-of-the-art model.