A new K-NN query algorithm based on the dynamic movement of spatial objects

• Published in: 2010 International Conference on Educational and Information Technology Vol. 1; pp. V1-376 - V1-379
• Main Authors: Guobin Li, Jine Tang
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.09.2010
• Subjects: Accuracy; dynamic movement; Geographic information systems; grid division; Heuristic algorithms; Indexes; Information technology; K-NN query algorithm; Nearest neighbor methods; relative position; Search problems; Sorting; Spatial databases; Spatial indexes
• ISBN: 1424480337; 9781424480333
• DOI: 10.1109/ICEIT.2010.5607687

The K-NN query algorithm is an important class of query algorithm in spatial database, the traditional K-NN query algorithm used the measurement distance and pruning strategy to search in the adopted index tree, the regional division algorithm based on the grid obtained the nearest neighbor objects by calculating the distance between the queried object and the data objects within the eight grids around the queried object, and carried on expanding inquiry scope outward layer by layer, in view of the advantage of the grid index, a new K-NN query algorithm based on the dynamic movement of spatial objects is proposed in this paper, it only carries on grid division to a small part of region around the queried object, when all the data objects in the current grid are already searched, make the cells in the grid empty, dynamically move the external data objects into the grid area continue to query according to the relative position of these objects, because the algorithm uses a fixed grid area to query the data objects in the whole region, so it can reduce the target storage space and the complexity degree of the algorithm calculation, the experiments show that the new query algorithm is superior to the traditional algorithms in the querying performance when the amount of data is increased, and the query efficiency is greatly increased.