Bitte benutzen Sie diese Referenz, um auf diese Ressource zu verweisen:
doi:10.22028/D291-24787
Dateien zu diesem Datensatz:
Datei | Beschreibung | Größe | Format | |
---|---|---|---|---|
RR_91_03.pdf | 15,98 MB | Adobe PDF | Öffnen/Anzeigen |
Titel: | Qualifying number restrictions in concept languages |
VerfasserIn: | Hollunder, Bernhard Baader, Franz |
Sprache: | Englisch |
Erscheinungsjahr: | 1991 |
Quelle: | Kaiserslautern ; Saarbrücken : DFKI, 1991 |
Kontrollierte Schlagwörter: | Künstliche Intelligenz Terminologische Sprache |
DDC-Sachgruppe: | 004 Informatik |
Dokumenttyp: | Forschungsbericht (Report zu Forschungsprojekten) |
Abstract: | We investigate the subsumption problem in logic-based knowledge representation languages of the KL-ONE family. The language presented in this paper provides the constructs for conjunction, disjunction, and negation of concepts, as well as qualifying number restrictions. The latter ones generalize the well-known role quantifications (such as value restrictions) and ordinary number restrictions, which are present in almost all KL-ONE based systems. Until now, only little attempts were made to integrate qualifying number restrictions into concept languages. It turns out that all known subsumption algorithms which try to handle these constructs are incomplete, and thus detecting only few subsumption relations between concepts. We present a subsumption algorithm for our language which is sound and complete. Subsequently we discuss why the subsumption problem in this language is rather hard from a computational point of view. This leads to an idea of how to recognize concepts which cause tractable problems. |
Link zu diesem Datensatz: | urn:nbn:de:bsz:291-scidok-35623 hdl:20.500.11880/24843 http://dx.doi.org/10.22028/D291-24787 |
Schriftenreihe: | Research report / Deutsches Forschungszentrum für Künstliche Intelligenz [ISSN 0946-008x] |
Band: | 91-03 |
Datum des Eintrags: | 11-Mär-2011 |
Fakultät: | SE - Sonstige Einrichtungen |
Fachrichtung: | SE - DFKI Deutsches Forschungszentrum für Künstliche Intelligenz |
Sammlung: | SciDok - Der Wissenschaftsserver der Universität des Saarlandes |
Alle Ressourcen in diesem Repository sind urheberrechtlich geschützt.