Reasoner = Logical Calculus + Rule Engine

Aus International Center for Computational Logic
Wechseln zu:Navigation, Suche

Toggle side column
Reasoner = Logical Calculus + Rule Engine

David Carral, Irina Dragoste, Markus Krötzsch
Reasoner = Logical Calculus + Rule Engine
KI, 2020
  • KurzfassungAbstract
    We propose using rule languages to encode complex reasoning algorithms in a declarative way. This approach -- which follows the classical slogan "Algorithm = Logic + Control" -- promises to turn high-level specifications of logical calculi as systems of inference rules into declarative rule-based models that can be executed on state-of-the-art rule engines.

    More precisely, given some input reasoning algorithm for some logic, we show how to produce a rule-based calculus; that is, a fixed rule set that can be used to replicate the consequences of the input algorithm. Simple rule languages suffice for simple logics, and we review our results on using Datalog rules to reason in the description logic EL. For more expressive logics, a suitably expressive yet implementable rule language often seems to be missing. To fill this gap, we consider an extension of Datalog with sets, Datalog(S), that can be executed by modern existential-rule reasoners, and we use it to present a rule-based reasoning calculus for the expressive description logic ALC.

    Because of some restrictions imposed by Springer, we cannot publish the manuscript on this page. To access the paper, click on the link below.
  • Weitere Informationen unter:Further Information: Link
  • Projekt:Project: CPEC
  • Forschungsgruppe:Research Group: Wissensbasierte SystemeKnowledge-Based Systems
The final publication is available at Springer via
  author    = {David Carral and Irina Dragoste and Markus Kr{\"{o}}tzsch},
  title     = {Reasoner = Logical Calculus + Rule Engine},
  journal   = {KI},
  publisher = {Springer},
  year      = {2020},
  doi       = {10.1007/s13218-020-00667-6}