Abstract:Collaborative filtering (CF) works have demonstrated the robust capabilities of Shallow Autoencoders on implicit feedback, showcasing highly competitive performance with other reasonable approaches (e.g., iALS and VAE-CF). However, despite their dual advantages of high performance and simple construction, EASE still exhibits several major shortcomings that must be addressed. To be more precise, the scalability of EASE is limited by the number of items, which determines the storage and inversion cost of a large dense matrix; the square-loss optimization objective does not consistently meet the recommendation task’s requirement for predicting personalized rankings, resulting in sub-optimal outcomes; the regularization coefficients are sensitive and require re-calibration with different datasets, leading to an exhaustive and time-consuming fine-tuning process. In order to address these obstacles, we propose a novel approach called Similarity-Structure Aware Shallow Autoencoder (AutoS 2 AE) that aims to enhance both recommendation accuracy and model efficiency. Our method introduces three similarity structures: Co-Occurrence, KNN, and NSW graphs, which replace the large dense matrix in EASE with a sparse structure, thus facilitating model compression. Additionally, we optimize the model by incorporating a low-rank training component into the matrix and applying a weighted square loss for improved ranking-oriented approximations. To automatically tune the hyperparameters, we further design two validation losses on the validation set for guidance and update the hyperparameters using the gradients of these validation losses. Both theoretical analyses regarding the introduction of similarity structures and empirical evaluations on multiple real-world datasets demonstrate the effectiveness of our proposed method, which significantly outperforms competing baselines.

An Efficient Method for Autoencoder‐based Collaborative Filtering

High-order autoencoder with data augmentation for collaborative filtering

Improving Recommendations With Collaborative Factors

Hybrid Collaborative Filtering with Autoencoders

Efficient Data-specific Model Search for Collaborative Filtering

Training Deep AutoEncoders for Collaborative Filtering

A Neural Autoregressive Approach to Collaborative Filtering

Neural Collaborative Autoencoder

Collaborative Autoencoder for Recommender Systems

Collaborative Filtering with User-Item Co-Autoregressive Models

Evolving Deep Neural Networks for Collaborative Filtering

SVD-AE: Simple Autoencoders for Collaborative Filtering

Deep Collaborative Autoencoder for Recommender Systems: A Unified Framework for Explicit and Implicit Feedback.

Attentive and Collaborative Deep Learning for Recommendation

Collaborative Filtering Based on Diffusion Models: Unveiling the Potential of High-Order Connectivity

A Dynamic Collaborative Filtering Algorithm based on Convolutional Neural Networks and Multi-layer Perceptron

A Neural Autoregressive Framework for Collaborative Filtering.

Deep Collaborative Filtering Incorporating Auxiliary Multi-Media Information.

Automated Sparse and Low-Rank Shallow Autoencoders for Recommendation

NeuSE: A Neural Snapshot Ensemble Method for Collaborative Filtering

Collaborative Reflection-Augmented Autoencoder Network for Recommender Systems