Towards a unified framework for efficient access methods and query operations in spatio-temporal databases

Spatio-temporal databases are required to efficiently support queries on large numbers of continuously moving objects. In this thesis we propose three techniques to address this challenge. We develop the STRIPES indexing method, which indexes predicted trajectories in a dual transformed space. Trajectories for objects in d-dimensional space are transformed into points in 2d-dimensional space and are indexed with a hierarchical grid decomposition structure. STRIPES can evaluate a range of queries including time-slice, window, and moving queries. Extensive experimental evaluation shows that STRIPES is significantly faster than leading existing predicted trajectory index (the TPR*-tree) for both updates and queries. The All Nearest Neighbor (ANN) operation is a commonly used primitive for analyzing large multi-dimensional datasets. Traditional R*-tree based methods use a pruning metric called MAXMAXDIST, which allows the algorithms to prune nodes in the index that need not be examined during the ANN computation. We introduce a new pruning metric called NXNDIST, and show that this metric is far more effective than MAXMAXDIST. We also propose the MBRQT index structure, and using extensive experimental evaluation show that MBRQT offers better speedup in ANN computation than the commonly used R*-tree index. In addition, we present the MBA algorithm using depth-first index traversal and bi-directional node expansion. Furthermore, we extend our method to evaluate the more general All-k-Nearest-Neighbor (AkNN) operation. Traditional spatial and traditional temporal joins have been widely studied in the past, but there is very little work on the more complex problem of trajectory joins. We present a general framework called JiST, that introduces a broad class of trajectory join operations and offers a set of algorithms to efficiently evaluate these operations. We then introduce the notion of Trajectory Privacy, and show the application of the JiST framework in the context of privacy preservation. Finally, we present detailed experimental results that demonstrate the efficiency and scalability of the JiST join algorithms. To the best of our knowledge, JiST is the first comprehensive framework for complex trajectory join operations and paves the foundation for building a complex querying platform for emerging trajectory based applications.