Enabling Spatial Queries in Open Government Data Portals

摘要

Recently, many governments have developed open government data portals as a way to facilitate the finding and the access to datasets produced by their agencies. The development of these portals has facilitated the retrieval of this kind of data, but they still have significant limitations. One drawback of current portals concerns the resolution of queries with spatial constraints. Many portals solve spatial queries selecting the datasets that contain in their description the place name informed by the user, which can lead to queries with low recall and precision. Aiming to solve these limitations, we propose a new spatial search engine to improve information retrieval in open government data portals. The main contributions of this work are the development of a system that retrieves OGD at the level of resources and the proposition of a ranking metric that evaluates the relevance of each resource retrieved from a query. We validated the proposed search engine using real data provided by the Brazilian open government data portal. The results obtained from the initial experiments showed that our solution is viable as it can retrieve data with good accuracy for many spatial queries of different granularities.