Multitemporal geospatial query grouping using correlation signatures

With recent advances in temporal and spatiotemporal databases, user demands are becoming more complex. As a result, simple queries are replaced by complex multitemporal query scenarios. In this paper we propose a novel image-based approach to temporally group together multidimensional geospatial queries. Correlation signatures act as a powerful raster mapping that visualizes multi-dimensional similarity of multiple queries and expresses it in a temporally referenced manner. Convolution of our raster representation with discrete weight masks can express arbitrary temporal preference (e.g. relative, cyclic queries). Furthermore, our multiquery grouping in the temporal domain can also support temporal relations between queries (e.g. alternative scenarios, AND/OR operators). By transforming the problem in the image domain, the expressiveness of our method allows an intuitive visual interaction to assist nonexpert database users.