Information-Enhanced Graph Neural Network for Transcending Homophily Barriers
Homophily and heterophily are intrinsic properties of graphs that describe whether linked nodes share similar properties. While Message Passing Neural Networks (MPNNs) have shown remarkable success in node classification tasks, their performance often deteriorates within specific homophily ranges, w...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10810421/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841536149306212352 |
|---|---|
| author | Xiao Liu Lijun Zhang Hui Guan |
| author_facet | Xiao Liu Lijun Zhang Hui Guan |
| author_sort | Xiao Liu |
| collection | DOAJ |
| description | Homophily and heterophily are intrinsic properties of graphs that describe whether linked nodes share similar properties. While Message Passing Neural Networks (MPNNs) have shown remarkable success in node classification tasks, their performance often deteriorates within specific homophily ranges, which we term the gray area. In this work, we identify and theoretically demonstrate the challenges faced by MPNNs in this gray area, highlighting the limitations of existing approaches in addressing it. To overcome these limitations, we propose the INformation-enhanced Graph Neural Network (INGNN), which introduces a novel framework that integrates three complementary features-ego-node features, graph structure features, and aggregated neighborhood features-through an adaptive feature fusion mechanism based on bi-level optimization. This design enables INGNN to transcend the homophily barriers and generalize effectively across the entire homophily spectrum. We validate the effectiveness of INGNN through extensive experiments on both synthetic and real-world datasets with different graph homophily. Specifically, INGNN outperforms 12 state-of-the-art MPNNs with an average rank of 1.78 on 9 real node classification datasets. Our ablation studies further show experimental evidence of how the integrated features contribute to the model’s performance under different homophily settings. INGNN is open-sourced and available at <uri>https://github.com/xl1990/ingnn</uri>. |
| format | Article |
| id | doaj-art-0407b8e5abe84ddf94278e174a498a35 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-0407b8e5abe84ddf94278e174a498a352025-01-15T00:01:36ZengIEEEIEEE Access2169-35362024-01-011219480419481510.1109/ACCESS.2024.352090310810421Information-Enhanced Graph Neural Network for Transcending Homophily BarriersXiao Liu0https://orcid.org/0000-0002-7888-6898Lijun Zhang1https://orcid.org/0000-0003-1946-7105Hui Guan2https://orcid.org/0000-0001-9128-2231College of Information and Computer Sciences, University of Massachusetts Amherst, Amherst, MA, USACollege of Information and Computer Sciences, University of Massachusetts Amherst, Amherst, MA, USACollege of Information and Computer Sciences, University of Massachusetts Amherst, Amherst, MA, USAHomophily and heterophily are intrinsic properties of graphs that describe whether linked nodes share similar properties. While Message Passing Neural Networks (MPNNs) have shown remarkable success in node classification tasks, their performance often deteriorates within specific homophily ranges, which we term the gray area. In this work, we identify and theoretically demonstrate the challenges faced by MPNNs in this gray area, highlighting the limitations of existing approaches in addressing it. To overcome these limitations, we propose the INformation-enhanced Graph Neural Network (INGNN), which introduces a novel framework that integrates three complementary features-ego-node features, graph structure features, and aggregated neighborhood features-through an adaptive feature fusion mechanism based on bi-level optimization. This design enables INGNN to transcend the homophily barriers and generalize effectively across the entire homophily spectrum. We validate the effectiveness of INGNN through extensive experiments on both synthetic and real-world datasets with different graph homophily. Specifically, INGNN outperforms 12 state-of-the-art MPNNs with an average rank of 1.78 on 9 real node classification datasets. Our ablation studies further show experimental evidence of how the integrated features contribute to the model’s performance under different homophily settings. INGNN is open-sourced and available at <uri>https://github.com/xl1990/ingnn</uri>.https://ieeexplore.ieee.org/document/10810421/Graph neural networks (GNNs)graph homophilygraph node classification |
| spellingShingle | Xiao Liu Lijun Zhang Hui Guan Information-Enhanced Graph Neural Network for Transcending Homophily Barriers IEEE Access Graph neural networks (GNNs) graph homophily graph node classification |
| title | Information-Enhanced Graph Neural Network for Transcending Homophily Barriers |
| title_full | Information-Enhanced Graph Neural Network for Transcending Homophily Barriers |
| title_fullStr | Information-Enhanced Graph Neural Network for Transcending Homophily Barriers |
| title_full_unstemmed | Information-Enhanced Graph Neural Network for Transcending Homophily Barriers |
| title_short | Information-Enhanced Graph Neural Network for Transcending Homophily Barriers |
| title_sort | information enhanced graph neural network for transcending homophily barriers |
| topic | Graph neural networks (GNNs) graph homophily graph node classification |
| url | https://ieeexplore.ieee.org/document/10810421/ |
| work_keys_str_mv | AT xiaoliu informationenhancedgraphneuralnetworkfortranscendinghomophilybarriers AT lijunzhang informationenhancedgraphneuralnetworkfortranscendinghomophilybarriers AT huiguan informationenhancedgraphneuralnetworkfortranscendinghomophilybarriers |