Data Pre-process Facilitating Efficient K-NN Queries in Spatial Database

EasyChair Preprint 2787

Abstract

This paper presents 2DMAX, a method to improve performance of large-scale repeated K-NN queries. Distances from a source point to its nearest neighbours are pre-computed and stored to facilitate re-use of query results for multiple queries without additional database access. For certain application scenarios this method may offer performance improvement up to one magnitude over conventional methods.

Keyphrases: Big Data Process, k-NN, nearest neighbour query, spatial database, spatial query

@booklet{EasyChair:2787,
  author    = {Sheng Zhou and Jonathan Simmons},
  title     = {Data Pre-process Facilitating Efficient K-NN Queries in Spatial Database},
  howpublished = {EasyChair Preprint 2787},
  year      = {EasyChair, 2020}}