Geospatial information and technologies have become a ubiquitous part of everyday services and is central to the business models of many of the digital disruptors that have become prominent in the 2010s. The rise of smartphones, tablets, and other mobile devices has contributed significantly to people’s expectation of the use of geospatial applications. User demand for increasing accuracy, currency, and detail is growing and processing will require more automated data capture and feature extraction to keep pace with those requirements.
GeoQA addresses these demands as it is focusing both on the representation and the efficient and accurate retrieval of geospatial knowledge. It creates the largest geospatial knowledge graph and develops novel technologies for its automatic update. It is the richest knowledge graph in terms of geospatial knowledge for the European Union and USA. It integrates data thematic and spatial attributes as linked data from: YAGO2; geographical administrative data provided by official sources of Greece, the United Kingdom, the Republic of Ireland and USA; the Global Administrative Areas dataset and OpenStreetMap.
At the same time, question answering systems are developed to deliver real-time and reliable knowledge-on-demand. One of our major goals is to answer questions of diverse forms that also require multiple resources for their answer. GeoQA systems are based on state-of-the-art technologies from the research areas of Natural Language Processing, Knowledge Graph Embeddings, Semantic Web and Linked Geospatial Data. Existing implementations of our team are further extended while at the same time also novel technologies are explored.