Michael Auer, Melanie Eckle, Sascha Fendrich, Fabian Kowatsch, Sabrina Marx, Martin Raifer, Moritz Schott, Rafael Troilo, Alexander Zipf, State of the Map 2018
Heidelberg University, Heidelberg Institute for Geoinformation Technology, GIScience Research Group, Department of Geography
[ Ссылка ]
Increasing use of OpenStreetMap (OSM) data in various applications and use cases lead to a growing number of research studies in which OSM data, its contributors and usage, and the quality of OSM data are analyzed. These studies include extrinsic and intrinsic data analyses, and provide interesting results for and about the community and data users. Such studies were mostly limited to analyzing either small samples of the OSM database or to simple types of analyses due to the capabilities of existing services and software operating on OSM’s full history data, which includes all data and every change made on a global scale. OSM data and contributor monitoring over longer timeframes and covering larger areas could however provide more comprehensive insights about our community and the data we are producing, also including spatial variations and evolution over time. In order to better monitor and understand OSM mapping, mappers and the produced data quality, we developed a software platform that applies big data technology to OSM full history data. This framework allows detailed analyses of the OSM data evolution and the detection of remarkable patterns over time. The setup of the framework, including openly accessible APIs, will allow interested community members and researchers of varying levels of experience to perform analyses with different levels of complexity. Initial analyses that have been conducted using the framework focused on user activity monitoring, addressing the following questions: Are we retaining new mappers? How many people participated in a specific area over time? How many new mappers have been active in contrast to experienced mappers? This kind of analyses enable the detection of user activity patterns in different regions over time. Results provide an overview of less and more actively mapped places in the OSM database. As all changes are considered, temporal patterns can be detected as well as changes due to individual users´ activity. Such insights help to understand user behaviors and to get a better sense of areas that are not well represented in the OSM database and community. These analyses are initial examples that show the potential of our framework. Further analyses will be discussed and addressed in collaboration with the OSM community and interested researchers. The talk will also be complemented by a workshop to enable attendees to learn how to make use of the framework for their own research questions and analyses.
