This paper introduces a formal model of a corpus of ontological knowledge in the domain of rubber compounds formulation. The model is based on a logical description of the entities living in the Chemical Formulation Problem, and its aim is to provides a set of necessary conditions that the entities must respect in order to be "ontologically consistent". Once the Chemical Formulation Problem has been interpreted and formalized within the AI paradigm of the Search, the formal representation of ontological knowledge have been exploited in order to drastically reduce the computational complexity of the resulting state space. A specific and domain dependent experimental campaign has shown that the efficiency of several search algorithms (i.e. A*, Iterative Deepening A*, Iterative Expansion and Branch and Bound) can be considerably improved by taking into account expert ontological knowledge about chemical compounds, and the formalization of this knowledge has led to the implementation of a automatic problem-solving system devoted to rubber compound formulation in truck tire industry.
- Mereological knowledge representation for the chemical formulation
- S BandiniAlessandro Mosca
- 2nd workshop "FOMI 2006" Formal Ontologies Meet Industry; University of Trento - ITALY; December 14-15, 2006, pp.55-69
- 2nd International Workshop on Formal Ontologies Meet Industry (FOMI 2006) (Trento, 14/12/2006–15/12/2006)
- Institute of Cognitive Sciences and Technologies, ISTC-CNR
- (UNIBZ)36444054
991007376572701241 - n.a.
- Faculty of Computer Science
- English
- Conference proceeding
- Bandini S, Mosca A
- description: Record is part of a bulk validation set