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.

The widespread use of mobile devices has resulted in the generation of vast amounts of spatial data. The availability of such large-scale spatial data facilitates the development of data-driven approaches to address real-life problems. This paper introduces the max coverage region (MCR) problem in road networks and provides efficient solutions. Given a set of spatial objects and a coverage radius, the MCR problem aims to identify a location from the road network, so that we can reach as many spatial objects as possible within the given coverage radius from the location. This problem is fundamental to supporting many real-world applications. Given a road network and a set of sensors, this problem can be used to find the best location for a sensor maintenance station. This problem can also be applied in medical research, such as in a protein–protein interaction network, where the nodes represent proteins, the edges represent their interactions, and the weight of an edge represents confidence. We can use the MCR problem to find the set of interacting proteins with a confidence budget. We propose an efficient exact solution to solve the problem, where we reduce the MCR problem to an equivalent problem named the most overlapped interval and design an edge-level upper bound estimation method to reduce the search space. Furthermore, we propose two approximate solutions that sacrifice a little accuracy for much better efficiency. Our experimental study on real-road network datasets demonstrates the effectiveness and superiority of the proposed approaches.

Self-supervised learning enables networks to learn discriminative features from massive data itself. Most state-of-the-art methods maximize the similarity between two augmentations of one image based on contrastive learning. By utilizing the consistency of two augmentations, the burden of manual annotations can be freed. Contrastive learning exploits instance-level information to learn robust features. However, the learned information is probably confined to different views of the same instance. In this paper, we attempt to leverage the similarity between two distinct images to boost representation in self-supervised learning. In contrast to instance-level information, the similarity between two distinct images may provide more useful information. Besides, we analyze the relation between similarity loss and feature-level cross-entropy loss. These two losses are essential for most deep learning methods. However, the relation between these two losses is not clear. Similarity loss helps obtain instance-level representation, while feature-level cross-entropy loss helps mine the similarity between two distinct images. We provide theoretical analyses and experiments to show that a suitable combination of these two losses can get state-of-the-art results. Code is available at https://github.com/guijiejie/ICCL.