# Detecting Beneficial Feature Interactions for Recommender Systems

arXiv: https://arxiv.org/abs/2008.00404 · HTML: https://mlsanity.com/paper/2008.00404/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Aug 2020.

## Built on by (most cited first, 22 of 22)

- [When Large Language Models Meet Personalization: Perspectives of Challenges and Opportunities](https://mlsanity.com/paper/2307.16376/) (2023)
- [MISSRec: Pre-training and Transferring Multi-modal Interest-aware Sequence Representation for Recommendation](https://mlsanity.com/paper/2308.11175/) (2023)
- [Mosaic of Modalities: A Comprehensive Benchmark for Multimodal Graph Learning](https://mlsanity.com/paper/2406.16321/) (2024)
- [MediTab: Scaling Medical Tabular Data Predictors via Data Consolidation, Enrichment, and Refinement](https://mlsanity.com/paper/2305.12081/) (2023)
- [Graph Pooling via Coarsened Graph Infomax](https://mlsanity.com/paper/2105.01275/) (2021)
- [GraphFM: Graph Factorization Machines for Feature Interaction Modeling](https://mlsanity.com/paper/2105.11866/) (2021)
- [WGCN: Graph Convolutional Networks with Weighted Structural Features](https://mlsanity.com/paper/2104.14060/) (2021)
- [Memorize, Factorize, or be Na"ive: Learning Optimal Feature Interaction Methods for CTR Prediction](https://mlsanity.com/paper/2108.01265/) (2021)
- [Adaptive Low-Precision Training for Embeddings in Click-Through Rate Prediction](https://mlsanity.com/paper/2212.05735/) (2022)
- [Neural Graph Matching based Collaborative Filtering](https://mlsanity.com/paper/2105.04067/) (2021)
- [AutoAlign: Fully Automatic and Effective Knowledge Graph Alignment enabled by Large Language Models](https://mlsanity.com/paper/2307.11772/) (2023)
- [Multi-domain Recommendation with Embedding Disentangling and Domain Alignment](https://mlsanity.com/paper/2308.05508/) (2023)
- [Detecting Arbitrary Order Beneficial Feature Interactions for Recommender Systems](https://mlsanity.com/paper/2206.13764/) (2022)
- [A Comprehensive Survey on Automated Machine Learning for Recommendations](https://mlsanity.com/paper/2204.01390/) (2022)
- [Ad-Rec: Advanced Feature Interactions to Address Covariate-Shifts in Recommendation Networks](https://mlsanity.com/paper/2308.14902/) (2023)
- [Decoupling and Damping: Structurally-Regularized Gradient Matching for Multimodal Graph Condensation](https://mlsanity.com/paper/2511.20222/) (2025)
- [FINT: Field-aware INTeraction Neural Network For CTR Prediction](https://mlsanity.com/paper/2107.01999/) (2021)
- [PHN: Parallel heterogeneous network with soft gating for CTR prediction](https://mlsanity.com/paper/2206.09184/) (2022)
- [LFGCF: Light Folksonomy Graph Collaborative Filtering for Tag-Aware Recommendation](https://mlsanity.com/paper/2208.03454/) (2022)
- [Beyond KAN: Introducing KarSein for Adaptive High-Order Feature Interaction Modeling in CTR Prediction](https://mlsanity.com/paper/2408.08713/) (2024)
- [DNN2LR: Automatic Feature Crossing for Credit Scoring](https://mlsanity.com/paper/2102.12036/) (2021)
- [Intrinsic and Extrinsic Factor Disentanglement for Recommendation in Various Context Scenarios](https://mlsanity.com/paper/2503.03524/) (2025)

## Abstract (arXiv)

Feature interactions are essential for achieving high accuracy in recommender systems. Many studies take into account the interaction between every pair of features. However, this is suboptimal because some feature interactions may not be that relevant to the recommendation result, and taking them into account may introduce noise and decrease recommendation accuracy. To make the best out of feature interactions, we propose a graph neural network approach to effectively model them, together with a novel technique to automatically detect those feature interactions that are beneficial in terms of recommendation accuracy. The automatic feature interaction detection is achieved via edge prediction with an L0 activation regularization. Our proposed model is proved to be effective through the information bottleneck principle and statistical interaction theory. Experimental results show that our model (i) outperforms existing baselines in terms of accuracy, and (ii) automatically identifies beneficial feature interactions.