Grass&Green is a project aims to improve the classification quality of OSM. Currently, we concern with grass-related classes. Every piece of land covered by grass has a precious value. However, the good classification of these entities is missing, due to unfamiliarity of the appropriate characteristics of each class. Agriculture lands, forest areas, urban green (e.g., park, recreation_ground,…etc.), villages and fields, meadow, and grass are all plausible classifications for a given entity. However, each class has specific characteristics that make one (or two) classification(s) is more appropriate than others.

In particular, the project focuses on grass entities within a city boundary in Germany. The project works to develop homogeneous classified entities. Based on the similarity with identical entities, the project guides the contributors toward appropriate classification. The project is part of research developed at University of Bremen by Ahmed Loai Ali.

To read more information about the research check the following articles:

• Data Quality Assurance for Volunteered Geographic Information here

• Ambiguity and Plausibility: Managing Classification Quality in Volunteered Geographic Information here

• Towards Rule-Guided Classification for Volunteered Geographic Information here

The project has a plan for further investigation on the classification of entities in VGI project, in particular OSM. However, currently we are developing our approach to learn the characteristics of specific classes, and hence guide the contributors toward appropriate classes.

Whatever your experience or mapping knowledge, your participation would help us to better understand the data. Now! enjoy the classification challenge on Grass&Green

Please, don’t hesitate to contact us. All your comments are Welcome.