Abstract
Over the recent years, several proposals were made to enhance database systems with automated reasoning. In this article we analyze two such enhancements based on meta-interpretation. We consider on the one hand the theorem prover Satchmo, on the other hand the Alexander and Magic Set methods. Although they achieve different goals and are based on distinct reasoning paradigms, Satchmo and the Alexander or Magic Set methods can be similarly described by upside-down meta-interpreters, i.e., meta-interpreters implementing one reasoning principle in terms of the other. Upside-down meta-interpretation gives rise to simple and efficient implementations, but has not been investigated in the past. This article is devoted to studying this technique. We show that it permits one to inherit a search strategy from an inference engine, instead of implementing it, and to combine bottom-up and top-down reasoning. These properties yield an explanation for the efficiency of Satchmo and a justification for the unconventional approach to top-down reasoning of the Alexander and Magic Set methods.
Dokumententyp: | Konferenzbeitrag (Anderer) |
---|---|
Fakultät: | Mathematik, Informatik und Statistik > Informatik |
Themengebiete: | 500 Naturwissenschaften und Mathematik > 510 Mathematik |
URN: | urn:nbn:de:bvb:19-epub-8509-2 |
Signatur: | print- |
Sprache: | Englisch |
Dokumenten ID: | 8509 |
Datum der Veröffentlichung auf Open Access LMU: | 17. Dez. 2008, 14:04 |
Letzte Änderungen: | 04. Nov. 2020, 12:51 |