Adapting propositional cases based on tableaux repairs using adaptation knowledge -- extended report

Gabin Personeni 1 Alice Hermann 1 Jean Lieber 1
1 ORPAILLEUR - Knowledge representation, reasonning
Inria Nancy - Grand Est, LORIA - NLPKD - Department of Natural Language Processing & Knowledge Discovery
Abstract : Adaptation is a step of case-based reasoning that aims at modifying a source case (representing a problem-solving episode) in order to solve a new problem, called the target case. An approach to adaptation consists in applying a belief revision operator that modifies minimally the source case so that it becomes consistent with the target case. Another approach consists in using domain-dependent adaptation rules. These two approaches can be combined: a revision operator parametrized by the adaptation rules is introduced and the corresponding revision-based adaptation uses the rules to modify the source case. This paper presents an algorithm for revision-based and rule-based adaptation based on tableaux repairs in propositional logic: when the conjunction of source and target cases is inconsistent, the tableaux method leads to a set of branches, each of them ending with clashes, and then, these clashes are repaired (thus modifying the source case), with the help of the adaptation rules. This algorithm has been implemented in the REVISOR/PLAK tool and some implementation issues are presented.
Document type :
Reports
Complete list of metadatas

Cited literature [19 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01011751
Contributor : Alice Hermann <>
Submitted on : Tuesday, June 24, 2014 - 2:57:48 PM
Last modification on : Friday, May 24, 2019 - 10:56:07 AM
Long-term archiving on : Wednesday, September 24, 2014 - 11:51:28 AM

File

report_on_revisor_plak.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01011751, version 1

Collections

Citation

Gabin Personeni, Alice Hermann, Jean Lieber. Adapting propositional cases based on tableaux repairs using adaptation knowledge -- extended report. [Research Report] Loria & Inria Grand Est. 2014. ⟨hal-01011751⟩

Share

Metrics

Record views

310

Files downloads

133