Point-of-Interest (POI) Data Validation Methods: An Urban Case Study
Open Access
- 29 October 2021
- journal article
- research article
- Published by MDPI AG in ISPRS International Journal of Geo-Information
- Vol. 10 (11) , 735
Abstract
Point-of-interest (POI) data from map sources are increasingly used in a wide range of applications, including real estate, land use, and transport planning. However, uncertainties in data quality arise from the fact that some of this data are crowdsourced and proprietary validation workflows lack transparency. Comparing data quality between POI sources without standardized validation metrics is a challenge. This study reviews and implements the available POI validation methods, working towards identifying a set of metrics that is applicable across datasets. Twenty-three validation methods were found and categorized. Most methods evaluated positional accuracy, while logical consistency and usability were the least represented. A subset of nine methods was implemented to assess four real-world POI datasets extracted for a highly urbanized neighborhood in Singapore. The datasets were found to have poor completeness with errors of commission and omission, although spatial errors were reasonably low (<60 m). Thematic accuracy in names and place types varied. The move towards standardized validation metrics depends on factors such as data availability for intrinsic or extrinsic methods, varying levels of detail across POI datasets, the influence of matching procedures, and the intended application of POI data.Keywords
This publication has 54 references indexed in Scilit:
- A Comprehensive Framework for Intrinsic OpenStreetMap Quality AnalysisTransactions in GIS, 2013
- Assessing Completeness and Spatial Error of Features in Volunteered Geographic InformationISPRS International Journal of Geo-Information, 2013
- Assessing the Effect of Data Imports on the Completeness of OpenStreetMap – A United States Case StudyTransactions in GIS, 2013
- IMPROVING VOLUNTEERED GEOGRAPHIC DATA QUALITY USING SEMANTIC SIMILARITY MEASUREMENTSThe International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2013
- Assessing similarity matching for possible integration of feature classifications of geospatial data from official and informal sourcesInternational Journal of Geographical Information Science, 2012
- How Many Volunteers Does it Take to Map an Area Well? The Validity of Linus’ Law to Volunteered Geographic InformationThe Cartographic Journal, 2010
- How Good is Volunteered Geographical Information? A Comparative Study of OpenStreetMap and Ordnance Survey DatasetsEnvironment and Planning B: Planning and Design, 2010
- Quality Assessment of the French OpenStreetMap DatasetTransactions in GIS, 2010
- QMatch – Using paths to match XML schemasData & Knowledge Engineering, 2007
- Bagging predictorsMachine Learning, 1996