A USER-DRIVEN SELECTION OF VGI BASED ON MINIMUM ACCEPTABLE QUALITY LEVELS

Abstract

Abstract. Despite Volunteered Geographic Information (VGI) activities are now extremely helpful in a number of scientific applications, researchers and decision makers oppose some resistance to the usage of volunteered contributions, due to quality issues. Several methods and workflows have been proposed to face quality issues in different VGI projects, usually built ad-hoc for specific datasets, thus resulting neither extensible nor transferable. In order to overcome this weakness, the authors propose to perform an user-driven assessment on VGI items in order to filter only those that satisfy minimally acceptable quality levels defined according to their specific quality requirements and project goals. In the present work the users, i.e., information consumers, are seen as decision makers and are allowed to set the minimum acceptable quality levels Thus the approach proposes a user driven assessment of the fitness for use of VGI items. The paper first briefly presents a view on VGI components and suitable quality indices, then it describes a logic architecture for managing them and for enabling a querying mechanism to the datasets. The approach is finally exemplified with a case study simulation.