More about HKUST
Spatial Indexes
PhD Qualifying Examination Title: "Spatial Indexes" by Mr. Moin Hussain MOTI Abstract: Spatial databases manage large datasets of objects with location information (e.g., GPS positions of smartphone users, vessel, and aircraft coordinates). Due to the sheer volume of data, and the absence of any natural ordering in multidimensional space, linear search for objects satisfying some spatial predicate (e.g., mobile users in the city center, the closest airplanes to the airport) is impractical. This motivates the need for spatial indexes that can efficiently filter and retrieve information. Spatial indexes are usually trees that either partition the space, or the data objects. The deepest tree level comprises leaf nodes, whereas the rest, including the root, consist of branches. Each node is associated with a spatial extent (e.g., a minimum bounding rectangle in 2D space) which covers all the nodes and objects in its subtree. Spatial indexes reside either in memory, or on the disk. This report focuses on the disk-based indexes, where nodes are stored on the disk, with their capacity limited by the disk-page size. The ability to fully utilize this capacity is critical to an index's overall performance. While spatial indexes are used in a variety of contexts, like moving objects, spatio-temporal aggregation, and cloud-based partitioning of spatial data, we focus on the most common spatial data processing tasks, such as the range, and the k-NN queries. We implemented popular spatial indexes like the Quad-Trees, the KDB-Trees, various R-Tree packing schemes, the R*-Tree, and the Waffle, and conducted extensive experimental analysis to understand how their characteristics contribute to their performances. We discuss our evaluation in this report. Date: Thursday, 4 May 2023 Time: 10:00am - 12:00noon Venue: Room 4472 Lifts 25/26 Committee Members: Prof. Dimitris Papadias (Supervisor) Prof. Siu-Wing Cheng (Chairperson) Prof. Raymond Wong Prof. Ke Yi **** ALL are Welcome ****