Semantic conflation in GIScience: a systematic review

Abstract

ABSTRACT Manifold providers from a wide range of initiatives (private organizations, volunteered efforts, social media, etc.) offer enormous data amounts with geospatial characteristics. These efforts of many data providers entail multiple data scenarios and imply many viewpoints about the same feature, involving different representations, accuracy, models, vocabularies, etc. Various techniques or processes are employed to deal with these heterogeneity problems related to diverse data sources within the conflation research area. However, semantic conflation has not been addressed widely in the literature, unlike geometrical conflation. Hence, it is unclear what issues semantic conflation tries to solve and what activities, methods, metrics, and techniques have been used in existing GIScience investigations. In this article, we carry out a systematic review of approaches that focus on semantic aspects for geospatial data conflation. Besides, we analyze a wide selection of contributions following different criteria to depict a detailed semantic conflation status in GIScience. Our contributions are: (i) an overview of semantic conflation application domains, (ii) a characterization of semantic issues within these domains, (iii) the recognition of gaps and weaknesses of collected researches, and (iv) several open challenges and opportunities for next steps in this GIScience research area.