Spatial-Temporal Hypergraph Self-Supervised Learning for Crime Prediction

Abstract

Crime has become a major concern in many cities, which calls for the rising demand for timely predicting citywide crime occurrence. Accurate crime prediction results are vital for the beforehand decision-making of government to alleviate the increasing concern about the public safety. While many efforts have been devoted to proposing various spatial-temporal forecasting techniques to explore dependence across locations and time periods, most of them follow a supervised learning manner, which limits their spatial-temporal representation ability on sparse crime data. Inspired by the recent success in self-supervised learning, this work proposes a Spatial-Temporal Self-Supervised Hypergraph Learning framework (ST-HSL) to tackle the label scarcity issue in crime prediction. Specifically, we propose the cross-region hypergraph structure learning to encode region-wise crime dependency under the entire urban space. Furthermore, we design the dual-stage self-supervised learning paradigm, to not only jointly capture local-and global-level spatial-temporal crime patterns, but also supplement the sparse crime representation by augmenting region self-discrimination. We perform extensive experiments on two real-life crime datasets. Evaluation results show that our ST-HSL significantly outperforms state-of-theart baselines. Further analysis provides insights into the superiority of our ST-HSL method in the representation of spatialtemporal crime patterns. The implementation code is available at https://github.com/LZH-YS1998/STHSL.

Explainable Spatio-Temporal Graph Neural Networks

“…Following the approach adopted in recent literature [20,41], we generate the training and testing sets in a ratio of 7:1. In the training set, crime records from the last month are used for validation purposes.…”

“…So, we summarize how to construct spatial and temporal graphs as follows: (i) spatial graph (𝐴 (𝑠 ) ): Spatial graph represents the correlations between spatial units. For the two common types of spatio-temporal prediction, i.e., graph-based and grid-based [15], we can construct graphs using a thresholded Gaussian kernel [19] and considering neighboring regions as neighbors [20,41], respectively. (ii) temporal graph (𝐴 (𝑡 ) ): Temporal graph represents the correlations between temporal representations at different time steps.…”

“…where 𝛿 denotes the hyperparameter for threshold. For the crime prediction, we instead utilize the mean absolute error (MSE) loss following [20] and have the following loss:…”