Method of functioning of intelligent agents, designed to solve action planning problems based on ontological approach
| Autor: | Miroslava Vovk, Petro Pukach, Vasyl Lytvyn, Victoria Vysotska, Dmytro Ugryn |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Computer science
020209 energy Process ontology 0211 other engineering and technologies Energy Engineering and Power Technology 02 engineering and technology Ontology (information science) computer.software_genre Industrial and Manufacturing Engineering Intelligent agent Management of Technology and Innovation Component (UML) 021105 building & construction 0202 electrical engineering electronic engineering information engineering Electrical and Electronic Engineering Agent architecture business.industry Applied Mathematics Mechanical Engineering Ontology-based data integration Intelligent decision support system Computer Science Applications Knowledge base Control and Systems Engineering Artificial intelligence business computer |
| Zdroj: | Eastern-European Journal of Enterprise Technologies. 3:11-17 |
| ISSN: | 1729-4061 1729-3774 |
| DOI: | 10.15587/1729-4061.2017.103630 |
| Popis: | The problem of operation of intelligent agents of action planning with the use of ontological approach was studied. Operation of intelligent agents is possible based on the knowledge of the subject-area, in other words, the knowledge base is used. Ontologies became the standard of knowledge base. Therefore, there arises the problem of development of methods and means of operation of intelligent systems based on ontologies, in particular intelligent agents of action planning. The method of functioning of intelligent agents of action planning based on ontologies was developed. For this purpose, weights of importance of concepts and relationships were introduced to the structure of ontology. These weights are used for finding a path in the space of states. The space of states itself is built by using the language of requests to ontology. Optimization problem, which assigns the rational behavior of an intelligent agent, is two-criterial. To solve it, we chose the method of the main component, if objective functions may be evaluated, or the method of complex criterion, if these functions are impossible to evaluate. Dimensionality of the space of states depends on the completeness of the ontology, and behavior effectiveness of an intelligent agent depends on the relevance of ontology. With this aim, in the course of automated development of ontology, we developed a method for evaluation of reliability of information sources that are used for developing ontologies. As a result of the studies, it was found that this approach allows us to increase operational efficiency of intelligent agents, if the process of ontology development is relevant to the needs of a subject domain. The developed approach may serve as a base for constructing a unified methodology for development of intelligent agents of action planning if ontology of a subject domain is the central component of this software complex |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|