User Defined Indexing

• Type: Article in International Conference Proceedings Book
• Date: 2009

Authors

David Vaz (Author) Other; Vitor Santos Costa (Author) FCUP; Michel Ferreira (Author) FCUP

Conference proceedings International

Title: LOGIC PROGRAMMING

Pages: 372-386

25th International Conference on Logic Programming (ICLP)

Pasadena, CA, JUL 14-17, 2009

Indexing

ISI Web of Knowledge - 4 Citations

Scopus - 6 Citations

Scientific classification

FOS: Natural sciences > Computer and information sciences

Other information

Authenticus ID: P-003-R7C

DOI: 10.1007/978-3-642-02846-5_31

Abstract (EN): Logic programming provides an ideal framework for tackling complex data, such as the multi-dimensional vector-based data used to represent spatial databases. Unfortunately, the usefulness of logic programming systems if often hampered by the fact that most of these systems have to rely on a single unification-based mechanism as the only way to search in the database. While unification can usually take effective advantage of hash-based indexing, it is often the case that queries over more complex and structured data, such as the vectorial terms stored in spatial databases, cannot. We propose a new extension to Prolog indexing: User Defined Indexing (UDI). In this mechanism, the programmer may add extra information to Prolog indices so that only interesting fragments of the database will be selected. UDI provides a general extension of indexing, and can be used for both instantiated and constrained variables. As a test case, we demonstrate how UDI can be combined with a constraint system to provide an elegant and efficient mechanism to generate and execute range queries and spatial queries. Experimental evaluation shows that this mechanism can achieve orders of magnitude speedups on non-trivial datasets.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 15