UniSTOK: Uniform Inductive Spatio-Temporal Kriging
This work is significant for researchers and practitioners in spatio-temporal data analysis, particularly in transportation and environmental monitoring, where sensor data often suffers from heterogeneous missingness.
This paper addresses the challenge of heterogeneous missingness in spatio-temporal kriging by proposing UniSTOK, a plug-and-play framework. UniSTOK enhances existing inductive kriging models by using a dual-branch input and explicit missingness mask modulation, leading to consistent and significant improvements across multiple real-world datasets.
Spatio-temporal kriging aims to infer signals at unobserved locations from observed sensors and is critical to applications such as transportation and environmental monitoring. In practice, however, observed sensors themselves often exhibit heterogeneous missingness, forcing inductive kriging models to rely on crudely imputed inputs. This setting brings three key challenges: (1) it is unclear whether an value is a true signal or a missingness-induced artifact; (2) missingness is highly heterogeneous across sensors and time; (3) missing observations distort the local spatio-temporal structure. To address these issues, we propose Uniform Inductive Spatio-Temporal Kriging (UniSTOK), a plug-and-play framework that enhances existing inductive kriging backbones under missing observation. Our framework forms a dual-branch input consisting of the original observations and a jigsaw-augmented counterpart that synthesizes proxy signals only at missing entries. The two branches are then processed in parallel by a shared spatio-temporal backbone with explicit missingness mask modulation. Their outputs are finally adaptively fused via dual-channel attention. Experiments on multiple real-world datasets under diverse missing patterns demonstrate consistent and significant improvements.