An approach for multi-scale urban building data integration and enrichment through geometric matching and semantic web


MEMDUHOĞLU A., BAŞARANER A. M.

CARTOGRAPHY AND GEOGRAPHIC INFORMATION SCIENCE, vol.49, no.1, pp.1-17, 2022 (Journal Indexed in SSCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 49 Issue: 1
  • Publication Date: 2022
  • Doi Number: 10.1080/15230406.2021.1952108
  • Title of Journal : CARTOGRAPHY AND GEOGRAPHIC INFORMATION SCIENCE
  • Page Numbers: pp.1-17
  • Keywords: Geospatial data integration, geospatial data enrichment, geospatial semantic web, urban buildings, multiple scales, SPATIAL DATA, LINKED DATA, INFORMATION, ONTOLOGIES

Abstract

The advent of Web 2.0 has emerged abundant but often unstructured user-generated georeferenced data, such as those from Volunteered Geographic Information (VGI) initiatives. In many cases, these data can be considered as complementary to the authoritative geospatial data. With the increasing availability of multi-source geospatial data, the efforts for geospatial data integration have gained momentum, aiming at gathering maximum information to answer sophisticated questions that cannot be answered using a single data source. Although there are various approaches employed for this purpose with different degrees of success, semantic web methods and tools have not been tested sufficiently in this scope, particularly for multi-scale urban building data integration and enrichment. Attempting to fill this gap, in this study, multi-source and multi-scale urban building data were integrated with a geometric matching method based on the overlapping area, then a geospatial ontology was developed to define multi-scale representations and detailed cardinality relations of the building features. Finally, some features from the geospatial ontology were then linked to popular knowledge bases such as DBpedia and YAGO. For the exploitation on the web, query and visualization processes were demonstrated using sample questions. The semantic web enabled to model complex cardinality of relations between the features from three different building data sets using inferencing and Semantic Web Rule Language (SWRL). The study showed that integrating different geospatial data sets as a knowledge base can facilitate answering sophisticated questions from different users.