Sparse reconstruction with spatial structures to automatically determine neighbors

ABSTRACT

Previous research has tended to use a global threshold of proximity to determine neighbors, neglecting spatial heterogeneity. Flexible thresholds implemented by adaptive search radii methods account for either the spatial structures or the non-spatial similarities of objects, but few consider both. By combining the spatial and non-spatial information of objects, we propose a novel approach that can automatically determine the neighbors that are strongly related to the object of interest. We introduce the sparse reconstruction technique from the signal processing domain, which aims to remove trivial relationships in a dataset. We extend the sparse reconstruction model by assuring three principles in spatial data, including retention of the correlation of data in the non-spatial attribute domain, preservation of local dependencies in the spatial domain, and removal of trivial relationships. Extensive experiments, based on road network missing value imputation and building clustering, show that our approach can make better use of both spatial and non-spatial information than a simple addition of them.

KEYWORDS:

• Spatial database
• spatial data mining
• sparse coding
• spatial inference

Acknowledgments

The authors are grateful to the associate editor and the anonymous referees for their valuable comments and suggestions.

Data and codes availability statement

The data and codes that support the findings of this study are available in [figshare.com] with the identifiers (https://doi.org/10.6084/m9.figshare.12361487.v1). The taxi trajectory data in Beijing and Chongqing cannot be made publicly available due to third party restrictions. Nevertheless, mocked taxi trajectory data are provided at the link to show how the codes work.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed here.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [42071442,41701440]; Natural Science Foundation of Hubei Province [2018CFB513]; National Key Research and Development Program of China [2017YFB0503500]; Fundamental Research Funds for the Central Universities, China University of Geosciences (Wuhan) [CUG170640]; State Key Laboratory of Resources and Environmental Information System [201801].

Notes on contributors

Wenhao Yu

Wenhao Yu is an Associate Professor in the School of Geography and Information Engineering, China University of Geosciences, Wuhan, China. His research interests include spatial database, map generalization, and spatial data mining. He is an editorial board member of PLOS ONE.

Yifan Zhang

Yifan Zhang is a master student in the School of Geography and Information Engineering, China University of Geosciences, Wuhan, China. His research interests include map generalization and spatial analysis.

Zhanlong Chen

Zhanlong Chen is a Professor in the School of Geography and Information Engineering, China University of Geosciences, Wuhan, China. His research interests include spatio-temporal Big Data and high-performance spatial computing.

Tinghua Ai

Tinghua Ai is a Professor in the School of Resource and Environment Science, Wuhan University, Wuhan, China. His research interests include spatio-temporal data mining and map generalization.