The mqr-tree for Very Large Object Sets
Wendy Osborn, Marc Moreau · 2015
This paper presents an evaluation of the mqr-tree for indexing a database containing a very large number of objects. Many spatial access methods have been proposed for handling either point and/or region data, with the vast majority able to handle a limited number of instances of these data types efficiently. However, many established and emerging application areas, such as recommender systems, require the management and indexing of very large object sets, such as a million places of interest that are each represented with a point. Using between one and five million points and objects, a comparison of both index construction and spatial query evaluation is performed versus a benchmark spatial indexing strategy. We show that the mqr-tree achieves significantly lower overlap and overcoverage when used to index a very large collection of objects. Also, the mqr-tree achieves significantly improved query processing performance in many cases. Therefore, the mqr-tree is a significant candidate for handling very large object sets for emerging applications.