Representing Spatial Data with Graph Contrastive Learning
Large-scale geospatial data pave the way for geospatial machine learning algorithms, and a good representation is related to whether the machine learning model is effective. Hence, it is a critical task to learn effective feature representation for geospatial data. In this paper, we construct a spat...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/4/880 |
| _version_ | 1827755689107259392 |
|---|---|
| author | Lanting Fang Ze Kou Yulian Yang Tao Li |
| author_facet | Lanting Fang Ze Kou Yulian Yang Tao Li |
| author_sort | Lanting Fang |
| collection | DOAJ |
| description | Large-scale geospatial data pave the way for geospatial machine learning algorithms, and a good representation is related to whether the machine learning model is effective. Hence, it is a critical task to learn effective feature representation for geospatial data. In this paper, we construct a spatial graph from the locations and propose a geospatial graph contrastive learning method to learn the location representations. Firstly, we propose a skeleton graph in order to preserve the primary structure of the geospatial graph to solve the positioning bias problem of remote sensing. Then, we define a novel mixed node centrality measure and propose four data augmentation methods based on the measure. Finally, we propose a heterogeneous graph attention network to aggregate information from both the structural neighborhood and semantic neighborhood separately. Extensive experiments on both geospatial datasets and non-geospatial datasets are conducted to illustrate that the proposed method outperforms state-of-the-art baselines. |
| first_indexed | 2024-03-11T08:12:34Z |
| format | Article |
| id | doaj.art-20ff89b3267b4b4f8ddd429c99a45ccf |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-11T08:12:34Z |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-20ff89b3267b4b4f8ddd429c99a45ccf2023-11-16T23:00:45ZengMDPI AGRemote Sensing2072-42922023-02-0115488010.3390/rs15040880Representing Spatial Data with Graph Contrastive LearningLanting Fang0Ze Kou1Yulian Yang2Tao Li3School of Cyber Science and Engineering, Frontiers Science Center for Mobile Information Communication and Security, Southeast University, Nanjing 210096, ChinaSchool of Cyber Science and Engineering, Frontiers Science Center for Mobile Information Communication and Security, Southeast University, Nanjing 210096, ChinaSchool of Cyber Science and Engineering, Frontiers Science Center for Mobile Information Communication and Security, Southeast University, Nanjing 210096, ChinaSchool of Cyber Science and Engineering, Frontiers Science Center for Mobile Information Communication and Security, Southeast University, Nanjing 210096, ChinaLarge-scale geospatial data pave the way for geospatial machine learning algorithms, and a good representation is related to whether the machine learning model is effective. Hence, it is a critical task to learn effective feature representation for geospatial data. In this paper, we construct a spatial graph from the locations and propose a geospatial graph contrastive learning method to learn the location representations. Firstly, we propose a skeleton graph in order to preserve the primary structure of the geospatial graph to solve the positioning bias problem of remote sensing. Then, we define a novel mixed node centrality measure and propose four data augmentation methods based on the measure. Finally, we propose a heterogeneous graph attention network to aggregate information from both the structural neighborhood and semantic neighborhood separately. Extensive experiments on both geospatial datasets and non-geospatial datasets are conducted to illustrate that the proposed method outperforms state-of-the-art baselines.https://www.mdpi.com/2072-4292/15/4/880spatial datacontrastive learninggraph representationlocation prediction |
| spellingShingle | Lanting Fang Ze Kou Yulian Yang Tao Li Representing Spatial Data with Graph Contrastive Learning Remote Sensing spatial data contrastive learning graph representation location prediction |
| title | Representing Spatial Data with Graph Contrastive Learning |
| title_full | Representing Spatial Data with Graph Contrastive Learning |
| title_fullStr | Representing Spatial Data with Graph Contrastive Learning |
| title_full_unstemmed | Representing Spatial Data with Graph Contrastive Learning |
| title_short | Representing Spatial Data with Graph Contrastive Learning |
| title_sort | representing spatial data with graph contrastive learning |
| topic | spatial data contrastive learning graph representation location prediction |
| url | https://www.mdpi.com/2072-4292/15/4/880 |
| work_keys_str_mv | AT lantingfang representingspatialdatawithgraphcontrastivelearning AT zekou representingspatialdatawithgraphcontrastivelearning AT yulianyang representingspatialdatawithgraphcontrastivelearning AT taoli representingspatialdatawithgraphcontrastivelearning |