A georeferenced graph model for geospatial data matching by optimising measures of similarity across multiple scales

ABSTRACT

The growth of georeferenced data sources calls for advanced matching methods to improve the reliability of geospatial data processing, such as map conflation. Existing matching methods mainly focus on similarity measures at the entity scale or area scale. A measure that combines entity-scale and area-scale similarities can provide sound matching results under various circumstances. In this paper, we propose a georeferenced-graph model that integrates multiscale similarities for data matching. Specifically, a match of correspondent data objects is identified by the entity-scale measure under the constraint of the area-scale measure. Nodes in the proposed georeferenced graph model represent polygons by their centroids, whereas the links in the graph connect the nodes (i.e. centroids) according to pre-defined rules. Then, we develop an algorithm to identify many-to-many matches. We demonstrate the proposed graph model and algorithm in real-world experiments using OpenStreetMap data. The experimental results show that the proposed georeferenced-graph model can effectively integrate the context and the location-and-form distance of geospatial data matches across different datasets.

KEYWORDS:

• Geospatial data matching
• georeferenced graph
• multiscale similarity relationships
• object matching
• optimisation

Disclosure statement

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

Data and codes availability statement

The data and codes that support the findings of this study are available with a DOI at 10.6084/m9.figshare.11521389. Neither of the reference datasets can be made publicly available due to propriety protection.

Additional information

Funding

This work was supported by the National Science Foundation of China [41421001]; National Key Research and Development Program [2017YFB0503501]; National Natural Science Foundation for Distinguished Young Scholars of China [41725006]; Hong Kong Research Grants Council [CUHK 14406514].

Notes on contributors

Wen-Bin Zhang

Wen-Bin Zhang is a Doctoral Candidate in the Institute of Geographical Sciences and Natural Resources Research at Chinese Academy of Sciences, and in the College of Resources and Environment at the University of Chinese Academy of Sciences, Beijing, China. E-mail: [email protected]. His research focuses on the statistical approach to spatial analysis and GIScience.

Yong Ge

Yong Ge is currently a Professor of the State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, CAS. E-mail: [email protected]. Her research interest mainly focuses on spatiotemporal statistics.

Yee Leung

Yee Leung is Emeritus Professor in the Department of Geography and Resource Management and Senior Research Fellow in the Institute of Future Cities at The Chinese University of Hong Kong, Hong Kong, China. E-mail: [email protected]. His research focuses on the statistical approach to uncertainty analysis and propagation in GIS, fuzzy set approach to geographical analysis and planning, intelligent spatial decision support systems, artificial intelligence, spatial data mining and knowledge discovery, and remote sensing.

Yu Zhou

Yu Zhou is a Research Assistant Professor in the Institute of Future Cities at The Chinese University of Hong Kong, Hong Kong, China. E-mail: [email protected]. His research interests include applied mathematics, such as nonlinear dynamics, especially fractals, and time series analysis and geography, including spatial analysis, quantitative methods, and geocomputation.