Semantic Enhancement of Volunteered Geographic Information

Laura Di Rocco
2016 International Conference of the Italian Association for Artificial Intelligence  
The continue use of social media has created a revolution in the production of data in terms of heterogeneity and volume. This huge amount of data is often georeferenced, and the geospatial position of data is a very relevant dimension for analysis. The considerable heterogeneity and the variable quality of user-generated data prevent however the full exploitation of this information. In this PhD project, we focus on user-generated geospatial data, commonly referred to as Volunteered Geographic
more » ... Information, where geographical objects are annotated by using tags. The simplicity to use tags allows users to put a lot of tags on objects. This creates a big and noisy tagspace in which it is hard to find data tagged by other people. To overcome this heterogeneity and quality problem, a solution is to rely on ontologies to classify spatial entities tags and names. This problem is addressed in the PhD project with specific reference to the tags used in OpenStreetMap for describing georeferenced objects relevant in urban context, producing as result an ontology for classifying such objects, and a corresponding ontology population approach.
dblp:conf/aiia/Rocco16 fatcat:tz6yo7tobrb3jcgdyzxltk4m74