Knowledge Engineering Method Based on Consensual Knowledge and Trust Computation: The MUSCKA System

  • Fabien Amarger
  • Jean-Pierre Chanet
  • Ollivier Haemmerlé
  • Nathalie Hernandez
  • Catherine Roussey
Conference paper

DOI: 10.1007/978-3-319-40985-6_14

Part of the Lecture Notes in Computer Science book series (LNCS, volume 9717)
Cite this paper as:
Amarger F., Chanet JP., Haemmerlé O., Hernandez N., Roussey C. (2016) Knowledge Engineering Method Based on Consensual Knowledge and Trust Computation: The MUSCKA System. In: Haemmerlé O., Stapleton G., Faron Zucker C. (eds) Graph-Based Representation and Reasoning. ICCS 2016. Lecture Notes in Computer Science, vol 9717. Springer, Cham

Abstract

We propose a method for building a knowledge base addressing specific issues such as covering end-users’ needs. After designing an ontology module representing the knowledge needed, we enrich and populate it automatically with knowledge extracted from existing sources such as thesauri or classifications. The originality of our proposition is to propose ontological object candidates from existing sources according to their relatedness to the ontological module and to their trust score. This paper describes the trust measures we propose which are obtained by analysing the consensus found in existing sources. We consider that knowledge is more reliable if it has been extracted from several sources. Our measures has been evaluated on a real case study with experts from the agriculture domain.

Keywords

Ontology development Trust Non-ontological sources Ontology Design Pattern Ontology merging 

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  • Fabien Amarger
    • 1
    • 2
  • Jean-Pierre Chanet
    • 1
  • Ollivier Haemmerlé
    • 2
  • Nathalie Hernandez
    • 2
  • Catherine Roussey
    • 1
  1. 1.Irstea, UR TSCF Technologies et systèmes d’information pour les agrosystèmesAubiéreFrance
  2. 2.Département de Mathématiques-InformatiqueIRIT, UMR 5505, UT2JToulouse CedexFrance

Personalised recommendations