Lune

ICLR2025Top-tier venue

Let Your Features Tell The Differences: Understanding Graph Convolution By Feature Splitting

Yilun Zheng, Xiang Li, Sitao Luan, Xiaojiang Peng, Lihui Chen

Abstract

Graph Neural Networks (GNNs) have demonstrated strong capabilities in processing structured data. While traditional GNNs typically treat each feature dimension equally important during graph convolution, we raise an important question: Is the graph convolution operation equally beneficial for each feature? If not, the convolution operation on certain feature dimensions can possibly lead to harmful effects, even worse than convolution-free models. Therefore, it is required to distinguish convolution-favored and convolution-disfavored features. Traditional feature selection methods mainly focus on identifying informative features or reducing redundancy, but they are not suitable for structured data as they overlook graph structures. In graph community, some studies have investigated the performance of GNN with respect to node features using feature homophily metrics, which assess feature consistency across graph topology. Unfortunately, these metrics do not effectively align with GNN performance and cannot be reliably used for feature selection in GNNs. To address these limitations, we introduce a novel metric, Topological Feature Informativeness (TFI), to distinguish GNNfavored and GNN-disfavored features, where its effectiveness is validated through both theoretical analysis and empirical observations. Based on TFI, we propose a simple yet effective Graph Feature Selection (GFS) method, which processes GNN-favored and GNN-disfavored features with GNNs and non-GNN models separately. Compared to original GNNs, GFS significantly improves the extraction of useful topological information from each feature with comparable computational costs. Extensive experiments show that after applying GFS to 8 baseline and state-of-the-art (SOTA) GNN architectures across 10 datasets, 90% of the GFS-augmented cases show significant performance boosts. Furthermore, our proposed TFI metric outperforms other feature selection methods for GFS. These results verify the effectiveness of both GFS and TFI. Additionally, we demonstrate that GFS's improvements are robust to hyperparameter tuning, highlighting its potential as a universally valid method for enhancing various GNN architectures. To facilitate reproducibility and further research, we have made our code publicly available at https://github.com/KTTRCDL/graph-feature-selection.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

6 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Mitigating Homophily Disparity in Graph Anomaly Detection: A Scalable and Adaptive Approach cites Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs
  • Mitigating Homophily Disparity in Graph Anomaly Detection: A Scalable and Adaptive Approach cites Geom-GCN: Geometric Graph Convolutional Networks
  • Mitigating Homophily Disparity in Graph Anomaly Detection: A Scalable and Adaptive Approach cites Beyond Low-frequency Information in Graph Convolutional Networks
  • Mitigating Homophily Disparity in Graph Anomaly Detection: A Scalable and Adaptive Approach cites Towards Anomaly-Aware Pre-Training and Fine-Tuning for Graph Anomaly Detection
  • Mitigating Homophily Disparity in Graph Anomaly Detection: A Scalable and Adaptive Approach cites Revisiting Heterophily For Graph Neural Networks
  • Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs cites Geom-GCN: Geometric Graph Convolutional Networks
  • Beyond Low-frequency Information in Graph Convolutional Networks cites Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs
  • Beyond Low-frequency Information in Graph Convolutional Networks cites Geom-GCN: Geometric Graph Convolutional Networks
  • Graph homophily booster: Reimagining the role of discrete features in heterophilic graph learning cites Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs
  • Graph homophily booster: Reimagining the role of discrete features in heterophilic graph learning cites Geom-GCN: Geometric Graph Convolutional Networks
  • Graph homophily booster: Reimagining the role of discrete features in heterophilic graph learning cites Beyond Low-frequency Information in Graph Convolutional Networks
  • Graph homophily booster: Reimagining the role of discrete features in heterophilic graph learning cites Revisiting Heterophily For Graph Neural Networks
  • Towards Anomaly-Aware Pre-Training and Fine-Tuning for Graph Anomaly Detection cites Beyond Low-frequency Information in Graph Convolutional Networks
  • Towards Anomaly-Aware Pre-Training and Fine-Tuning for Graph Anomaly Detection cites Revisiting Heterophily For Graph Neural Networks
  • GraphTARIF: Linear Graph Transformer with Augmented Rank and Improved Focus cites Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs
  • Revisiting Heterophily For Graph Neural Networks cites Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs
  • Revisiting Heterophily For Graph Neural Networks cites Geom-GCN: Geometric Graph Convolutional Networks
  • Revisiting Heterophily For Graph Neural Networks cites Beyond Low-frequency Information in Graph Convolutional Networks
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers6

  1. Graph homophily booster: Reimagining the role of discrete features in heterophilic graph learningRuizhong Qiu, Ting-Wei Li, Gaotang Li, Hanghang TongICLR 2026 · 2 citations
  2. Fixed Aggregation Features Can Rival GNNsCelia Rubio-Madrigal, Rebekka BurkholzICML 2026 · 2 citations
  3. Mitigating Homophily Disparity in Graph Anomaly Detection: A Scalable and Adaptive ApproachYunhui Liu, Qizhuo Xie, Yinfeng Chen, Xudong Jin et al.WWW 2026
  4. Towards Anomaly-Aware Pre-Training and Fine-Tuning for Graph Anomaly DetectionYunhui Liu, Jiashun Cheng, Yiqing Lin, Qizhuo Xie et al.ICLR 2026
  5. GraphTARIF: Linear Graph Transformer with Augmented Rank and Improved FocusZhaolin Hu, Kun Li, Hehe Fan, Yi YangWWW 2026

Builds on17

  1. Beyond Homophily in Graph Neural Networks: Current Limitations and Effective DesignsJiong Zhu, Yujun Yan, Lingxiao Zhao, Mark Heimann et al.NeurIPS 2020 · 1,490 citations
  2. Geom-GCN: Geometric Graph Convolutional NetworksHongbin Pei, Bingzhe Wei, Kevin Chen-Chuan Chang, Yu Lei et al.ICLR 2020 · 1,445 citations
  3. Beyond Low-frequency Information in Graph Convolutional NetworksDeyu Bo, Xiao Wang, Chuan Shi, Huawei ShenAAAI 2021 · 773 citations
  4. PyTorch 2: Faster Machine Learning Through Dynamic Python Bytecode Transformation and Graph CompilationJason Ansel, Edward Z. Yang, Horace He, Natalia Gimelshein et al.ASPLOS 2024 · 693 citations
  5. Revisiting Heterophily For Graph Neural NetworksSitao Luan, Chenqing Hua, Qincheng Lu, Jiaqi Zhu et al.NeurIPS 2022 · 351 citations

Related papers

  1. Characterizing Graph Datasets for Node Classification: Homophily-Heterophily Dichotomy and BeyondOleg Platonov, Denis Kuznedelev, Artem Babenko, Liudmila ProkhorenkovaNeurIPS 2023 · 95 citations
  2. Adaptive Node Feature Selection for Graph Neural NetworksMadeline Navarro, Ali Azizpour, Santiago SegarraICML 2026
  3. What Is Missing For Graph Homophily? Disentangling Graph Homophily For Graph Neural NetworksYilun Zheng, Sitao Luan, Lihui ChenNeurIPS 2024 · 24 citations
  4. GBK-GNN: Gated Bi-Kernel Graph Neural Networks for Modeling Both Homophily and HeterophilyLun Du, Xiaozhou Shi, Qiang Fu, Xiaojun Ma et al.WWW 2022 · 155 citations
  5. Boosting Graph Convolution with Disparity-induced Structural RefinementSujia Huang, Yueyang Pi, Tong Zhang, Wenzhe Liu et al.WWW 2025 · 1 citation
Dusk over the sea between two cliffs drawn in fine vertical lines