Estimating quality aspects of OpenStreetMap

Completeness refers to the relationship between the entities in the database and the ‘abstract universe’ of all such entities. ‘Model completeness’ refers to the agreement between the database specification and the abstract universe that is required for a particular database application (Brassel et al 1995). Model completeness (as opposite to ‘Data completeness’) is application-dependent and therefore an aspect of fitness-for-use. Intrinsic completeness methods try to estimate the completeness by trying to find a correlation to other entities in the database. Intrinsic completeness methods don't require external reference data.

The project "osmcross" investigates the possibilities to roughly estimate completenes on OSM using intrinsic methods combined with modern machine learning methods. The industry partner of this project is Data Ahead Analytics and it's partially funded by Innosuisse.

Do you have questions about OpenStreetMap, or more generally about open data, open source, open database systems or data analytics?

Contact: Stefan Keller

Here's a figure for osmcross with caption text:
Fig. Visualization of the prediction of completeness of gastronomy objects in OpenStreetMap in Bern based on training from Zurich (input tags are building, street, bus_station, railway, shop; coordinate reference system EPSG:3857)