An engine for computing well-founded models

Terrance Swift

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Scopus citations


The seemingly simple choice of whether to use call variance or call subsumption in a tabled evaluation deeply affects an evaluation's properties. Most tabling implementations have supported only call variance or, in the case of XSB Prolog, supported call subsumption only for stratified programs. However, call subsumption has proven critical for (sub-)model generation as required for some kinds of program analysis (e.g. type analysis) and for semantic web applications such as RDF inference. At the same time, the lack of well-founded negation has prevented the use of call subsumption in producing residual programs, and has limited its use in semantic web applications that require negation (e.g. evaluation of OWL ontologies). This paper describes an engine for evaluating normal programs under the well-founded semantics (WFS) in which the evaluation method can be based on a mixture of call subsumption and call variance, chosen at the predicate level. The implementation has been thoroughly tested for both local and batched evaluation and is available in version 3.2 of XSB.

Original languageEnglish (US)
Title of host publicationLogic Programming - 25th International Conference, ICLP 2009, Proceedings
Number of pages5
StatePublished - 2009
Event25th International Conference on Logic Programming, ICLP 2009 - Pasadena, CA, United States
Duration: Jul 14 2009Jul 17 2009

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5649 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other25th International Conference on Logic Programming, ICLP 2009
Country/TerritoryUnited States
CityPasadena, CA


  • Tabling
  • WAM

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'An engine for computing well-founded models'. Together they form a unique fingerprint.

Cite this