Improving Case Retrieval Using Typicality

Emmanuelle Gaillard 1, 2 Jean Lieber 1 Emmanuel Nauer 1
1 ORPAILLEUR - Knowledge representation, reasonning
Inria Nancy - Grand Est, LORIA - NLPKD - Department of Natural Language Processing & Knowledge Discovery
Abstract : This paper shows how typicality can be used to improve the case retrieval of a case-based reasoning (CBR) system, improving at the same time the global results of the CBR system. Typicality discriminates subclasses of a class in the domain ontology depending of how a subclass is a good example for its class. Our approach proposes to partition the subclasses of some classes into atypical, normal and typical subclasses in order to refine the domain ontology. The refined ontology allows a finer-grained generalization of the query during the retrieval process. The benefits of this approach are presented according to an evaluation in the context of Taaable, a \cbr system designed for the cooking domain.
Document type :
Conference papers
Complete list of metadatas

https://hal.inria.fr/hal-01178317
Contributor : Emmanuel Nauer <>
Submitted on : Saturday, July 18, 2015 - 5:26:42 PM
Last modification on : Friday, May 24, 2019 - 10:58:04 AM

Identifiers

  • HAL Id : hal-01178317, version 1

Collections

Citation

Emmanuelle Gaillard, Jean Lieber, Emmanuel Nauer. Improving Case Retrieval Using Typicality. 23rd International Conference on Case-Based Reasoning (ICCBR 2015), Sep 2015, Frankfurt am Main, Germany. ⟨hal-01178317⟩

Share

Metrics

Record views

343