Fuzzy spatio-temporal ontologies and formal construction based on fuzzy Petri nets.
Haitao Cheng,Li Yan,Zongmin Ma,Slobodan Ribaric
DOI: https://doi.org/10.1111/coin.12199
2019-01-01
Computational Intelligence
Abstract:Fuzzy spatio-temporal knowledge is required in a wide range of application fields, including GIS, spatio-temporal database, and artificial intelligence. Ontology, as a formal method of knowledge representation, plays a key role in the Semantic Web. Therefore, how to extend ontologies to represent fuzzy spatio-temporal knowledge needs to be solved. In this paper, we propose an approach for representing fuzzy spatio-temporal knowledge with ontologies and investigate a formal construction of fuzzy spatio-temporal ontologies based on fuzzy Petri nets. First, we propose a formal definition of fuzzy spatio-temporal ontologies. In addition, based on fuzzy description logic, a description logic named the fuzzy spatio-temporal description logic, which provides a logical basis for the fuzzy spatio-temporal ontologies, is presented. Furthermore, a fuzzy spatio-temporal web ontology language, which is fuzzy and spatio-temporal extension of standard ontology language OWL, is proposed to make the fuzzy spatio-temporal ontologies easy to use and support efficient reasoning. On the basis of fuzzy spatio-temporal ontologies, we investigate a formal approach for constructing fuzzy spatio-temporal ontologies from fuzzy spatio-temporal Petri nets, ie, transforming fuzzy spatio-temporal Petri nets (including Petri net model and Petri net instance) into fuzzy spatio-temporal ontologies at both structure and instance levels. Finally, we prove the correctness of the transformation and provide a detailed transformation example.