This year AGILE celebrated its 20th birthday and conference from May 10 – 12 at Wageningen University, Netherlands. The conference organizers chose “societal geo-information” to be the main theme of the research presented. The GIScience Research Group Heidelberg was represented by its members Tessio Novack, Franz-Benjamin Mocnik and Benjamin Herfort.
On Tuesday, the day before the opening of the conference, several workshops were held. The VGI-Analytics workshop, which was co-organized by the GIScience Research Group Heidelberg ,tackled current research topics related to Volunteered Geographic Information (VGI), in particular the integration of different VGI data sources and data analysis. Six (short) papers were presented, amongst them a paper about the inevitability of calibration in VGI quality assessment by Franz-Benjamin Mocnik. The participants even discussed in break-out groups legal issues of VGI, data quality aspects of VGI, as well as current technologic challenges. Complementing the workshop there is still an open call for journal publictions in the international open access journal 2 Geo-spatial Information Science (GSIS)” (Taylof & Francis) free of charge. The extended final Submission Deadline is 9th June 2017.
Benjamin Herfort presented our recent work on the analysis of MapSwipe data and development of the MapSwipe Analytics platform. The talk highlighted how intrinsic parameters can help to understand VGI data better. For MapSwipe data, especially the agreement among volunteers is a good predictor of correct or wrong crowdsourced classifications. In the next weeks we will continue our activities to improve the MapSwipe App and to develop methods to improve the overall data quality. Stay tuned and have a look at http://mapswipe.geog.uni-heidelberg.de.
Tessio Novack presented an analysis and method to match point of interest from different VGI sources based on work in the corresponding DFG project . The matching of features across different VGI projects may serve to assess and improve the reliability and completeness of VGI data. We propose a matching strategy based on a graph using several combinations of spatial and semantic similarities and string similarities and evaluate the quality of the different combinations.
Franz-Benjamin Mocnik presented a poster about how data quality and fitness for purpose relate. While both concepts are well known to be mutually dependent, they have not yet been properly defined in common terms. The poster presents an approach to define data qualit in terms of fitness for purpose, which renders possible to translate between them and to derive data quality by the computation of fitnesses for purpose. The work is part of a DFG project, with the aim to calibrate intrinsic quality assessment of VGI and OSM in particular.
We are looking forward to the next AGILE conference at Lund University in 2018. 🙂
References:
Herfort, B., Reinmuth, M., Porto de Albuquerque, J., Zipf, A. (2017): Towards evaluating the mobile crowdsourcing of geographic information about human settlements. 20th AGILE conference 2017, Wageningen, Netherlands.
Novack, T., Peters, R., Zipf, A. (2017): A graph-based strategy for matching points-ofinterests from different VGI sources. 20th AGILE conference 2017, Wageningen, Netherlands.
Mocnik, F.-B., Zipf, A., Fan, H. (2017): The Inevitability of Calibration in VGI Quality Assessment. 4th Workshop on Volunteered Geographic Information: Integration, Analysis, and Applications (VGI-Analytics), Wageningen, Netherlands.
Mocnik, F.-B., Zipf, A., Fan, H. (2017): Data Quality and Fitness for Purpose. Poster. 20th AGILE Conference on Geographic Information Science, Wageningen, Netherlands.