Heterogeneous Graph Purification Network: Purifying Noisy Heterogeneity without Metapaths
Heterogeneous graph neural networks (HGNNs) deliver the powerful capability to model many complex systems in real-world scenarios by embedding rich structural and semantic information of a heterogeneous graph into low-dimensional representations. However, existing HGNNs encounter great difficulty in...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-03-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/13/6/3989 |
| _version_ | 1797613591363321856 |
|---|---|
| author | Sirui Shen Daobin Zhang Shuchao Li Pengcheng Dong Qing Liu Xiaoyu Li Zequn Zhang |
| author_facet | Sirui Shen Daobin Zhang Shuchao Li Pengcheng Dong Qing Liu Xiaoyu Li Zequn Zhang |
| author_sort | Sirui Shen |
| collection | DOAJ |
| description | Heterogeneous graph neural networks (HGNNs) deliver the powerful capability to model many complex systems in real-world scenarios by embedding rich structural and semantic information of a heterogeneous graph into low-dimensional representations. However, existing HGNNs encounter great difficulty in balancing the ability to avoid artificial metapaths with resisting structural and informational noise in a heterogeneous graph. In this paper, we propose a novel framework called <b>H</b>eterogeneous <b>G</b>raph <b>P</b>urification <b>N</b>etwork (HGPN) which aims to solve such dilemma by adaptively purifying the noisy heterogeneity. Specifically, instead of relying on artificial metapaths, HGPN models heterogeneity by subgraph decomposition and adopts inter-subgraph and intra-subgraph aggregation methods. HGPN can learn to purify noisy edges based on semantic information with a parallel heterogeneous structure purification mechanism. Besides, we design a neighborhood-related dynamic residual update method, a type-specific normalization module and cluster-aware loss to help all types of node achieve high-quality representations and maintain feature distribution while preventing feature over-mixing problems. Extensive experiments are conducted on four common heterogeneous graph datasets, and results show that our approach outperforms all existing methods and achieves state-of-the-art performances consistently among all the datasets. |
| first_indexed | 2024-03-11T06:58:04Z |
| format | Article |
| id | doaj.art-e9794422ee144c0ab11f9369b8e80b2f |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-11T06:58:04Z |
| publishDate | 2023-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-e9794422ee144c0ab11f9369b8e80b2f2023-11-17T09:29:49ZengMDPI AGApplied Sciences2076-34172023-03-01136398910.3390/app13063989Heterogeneous Graph Purification Network: Purifying Noisy Heterogeneity without MetapathsSirui Shen0Daobin Zhang1Shuchao Li2Pengcheng Dong3Qing Liu4Xiaoyu Li5Zequn Zhang6Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, ChinaAerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, ChinaHeterogeneous graph neural networks (HGNNs) deliver the powerful capability to model many complex systems in real-world scenarios by embedding rich structural and semantic information of a heterogeneous graph into low-dimensional representations. However, existing HGNNs encounter great difficulty in balancing the ability to avoid artificial metapaths with resisting structural and informational noise in a heterogeneous graph. In this paper, we propose a novel framework called <b>H</b>eterogeneous <b>G</b>raph <b>P</b>urification <b>N</b>etwork (HGPN) which aims to solve such dilemma by adaptively purifying the noisy heterogeneity. Specifically, instead of relying on artificial metapaths, HGPN models heterogeneity by subgraph decomposition and adopts inter-subgraph and intra-subgraph aggregation methods. HGPN can learn to purify noisy edges based on semantic information with a parallel heterogeneous structure purification mechanism. Besides, we design a neighborhood-related dynamic residual update method, a type-specific normalization module and cluster-aware loss to help all types of node achieve high-quality representations and maintain feature distribution while preventing feature over-mixing problems. Extensive experiments are conducted on four common heterogeneous graph datasets, and results show that our approach outperforms all existing methods and achieves state-of-the-art performances consistently among all the datasets.https://www.mdpi.com/2076-3417/13/6/3989graph neural networkheterogeneous graph representation learningheterogeneous graph purificationgraph representation |
| spellingShingle | Sirui Shen Daobin Zhang Shuchao Li Pengcheng Dong Qing Liu Xiaoyu Li Zequn Zhang Heterogeneous Graph Purification Network: Purifying Noisy Heterogeneity without Metapaths Applied Sciences graph neural network heterogeneous graph representation learning heterogeneous graph purification graph representation |
| title | Heterogeneous Graph Purification Network: Purifying Noisy Heterogeneity without Metapaths |
| title_full | Heterogeneous Graph Purification Network: Purifying Noisy Heterogeneity without Metapaths |
| title_fullStr | Heterogeneous Graph Purification Network: Purifying Noisy Heterogeneity without Metapaths |
| title_full_unstemmed | Heterogeneous Graph Purification Network: Purifying Noisy Heterogeneity without Metapaths |
| title_short | Heterogeneous Graph Purification Network: Purifying Noisy Heterogeneity without Metapaths |
| title_sort | heterogeneous graph purification network purifying noisy heterogeneity without metapaths |
| topic | graph neural network heterogeneous graph representation learning heterogeneous graph purification graph representation |
| url | https://www.mdpi.com/2076-3417/13/6/3989 |
| work_keys_str_mv | AT siruishen heterogeneousgraphpurificationnetworkpurifyingnoisyheterogeneitywithoutmetapaths AT daobinzhang heterogeneousgraphpurificationnetworkpurifyingnoisyheterogeneitywithoutmetapaths AT shuchaoli heterogeneousgraphpurificationnetworkpurifyingnoisyheterogeneitywithoutmetapaths AT pengchengdong heterogeneousgraphpurificationnetworkpurifyingnoisyheterogeneitywithoutmetapaths AT qingliu heterogeneousgraphpurificationnetworkpurifyingnoisyheterogeneitywithoutmetapaths AT xiaoyuli heterogeneousgraphpurificationnetworkpurifyingnoisyheterogeneitywithoutmetapaths AT zequnzhang heterogeneousgraphpurificationnetworkpurifyingnoisyheterogeneitywithoutmetapaths |