Geographical Provenance of Open Government Datasets: Evaluating Geospatial Metadata in Municipal Open Data Portals

Abstract

The study investigates the hypothesis that the adoption of metadata standards such as DCAT in German municipal Open Data portals improves the assignability of datasets to specific geographical regions. It focuses on municipal Open Data portals and compares those that utilise DCAT-compliant metadata structures to portals with non-standard formats, such as ArcGIS JSON. The harvesting and analysis of metadata from each portal is conducted via public endpoints. The classification routine is based on metadata fields such as geocoding URI, dct:spatial and keywords. Each dataset is evaluated for its regional geographical capability. Preliminary results suggest that DCAT-based portals significantly outperform non-standard implementations in terms of regional assignability of datasets, error rate, and metadata completeness.