Neural Collaborative Filtering Vs. Matrix Factorization Revisited
2020 · Steffen Rendle, Walid Krichene, Li Zhang, et al.
Abstract
Embedding based models have been the state of the art in collaborative filtering for over a decade. Traditionally, the dot product or higher order equivalents have been used to combine two or more embeddings, e.g., most notably in matrix factorization. In recent years, it was suggested to replace the dot product with a learned similarity e.g. using a multilayer perceptron (MLP). This approach is often referred to as neural collaborative filtering (NCF). In this work, we revisit the experiments of the NCF paper that popularized learned similarities using MLPs. First, we show that with a proper hyperparameter selection, a simple dot product substantially outperforms the proposed learned similarities. Second, while a MLP can in theory approximate any function, we show that it is non-trivial to learn a dot product with an MLP. Finally, we discuss practical issues that arise when applying MLP based similarities and show that MLPs are too costly to use for item recommendation in production e
Authors
(none)
Tags
Stats
Related papers
- Predicting Pairwise Relations With Neural Similarity Encoders (2017)0.00
- Efficient Nearest Neighbor Search For Cross-encoder Models Using Matrix Factorization (2022)4.52
- Beyond Similarity: Relation Embedding With Dual Attentions For Item-based Recommendation (2019)0.00
- Embedding In Recommender Systems: A Survey (2023)0.00
- Retrieval With Learned Similarities (2024)0.00
- Speeding Up Memory-based Collaborative Filtering With Landmarks (2017)0.00
- Combigcn: An Effective GCN Model For Recommender System (2025)6.77
- Revisiting Neural Retrieval On Accelerators (2023)9.41