A hypercube-based data structure for spatio-temporal exploration and analysis

This paper examines the opportunities existing within the multidimensional database (MDDB) approach to support spatio-temporal data exploration and analysis (STEA). MDDB-specific technology used for Data Warehousing and On-Line Analytical Processing (OLAP) is recognized for supporting easy query building, rapid query run times and simple interactive navigation within multiple displays, including maps. However, due to the lack of topological operators (spatial and temporal) in MDDBs limited STEA can be achieved. We thus propose a new data structure which implements a spatiotemporal dimension of topological operators in multidimensional databases. It involves a hierarchy of topological operators which uses the spatial and temporal relationships between instances of object classes. This hierarchy is structured according to granularity levels (4 in our example). Operators at the highest level carry general meanings while operators from the lowest levels have very specific meanings. This hierarchy also takes into consideration behavior models of spatial analysis in order to improve database usability. The topological relationships are described for each level of the hierarchy, 2 either through groups of operators or through the ISO/TC 211 9-intersection method (9IM) developed by [1] and the DE+9IM developed by [2]. These topological relationships are computed a priori and stored in the MDDB structure, often called hypercube, as 9 digit strings compatible with the ISO/TC211 topological operators [3]. This proposed solution represents a new approach for rapid and intuitive spatio-temporal analysis while being generic, software-independent and implementable in today’s commercial MDDB engines.