Method for Ontology Content and Structure Optimization, Provided by a Weighted Conceptual Graph

• Published in: Webology Vol. 15; no. 2; pp. 66 - 85
• Main Authors: Lytvyn, Vasyl, Vysotska, Victoria, Dosyn, Dmytro, Burov, Yevhen
• Format: Journal Article
• Language: English
• Published: Tehran Dr. Alireza Noruzi, University of Tehran, Department of Library and Information Science 01.12.2018
• Subjects: Algorithms; Artificial intelligence; Automation; Control theory; Data mining; Fuzzy sets; International conferences; Linguistics; Methods; Ontology; Optimization techniques; Queries; Query expansion; Semantic web; Semantics; Taxonomy; Ukraine
• ISSN: 1735-188X

The complex dynamic structure of the ontology and its content requires implementing of optimization procedures to better ontology processing performance and resolve conflicts between data. This paper addresses the issues of ontology optimization with the purpose of adapting its content to the needs of users by excluding those items that are rarely used or not used at all, or which do not belong or not related to a particular subject area. The approach is based on automated weighting of concepts and relations during ontology learning. The ontology expanded in this way, is sequentially optimized according to criteria of integrity, absence of ambiguity, volume, response time, completeness and thematic balance. The optimization method of minimal spanning tree search was applied. This optimization task can be further reduced to the backpack problem for which the effective solving algorithms are known. The use of developed optimization techniques provides controlled automated ontology learning procedure that significantly expands the usability of ontology-based systems and reduces the time expenditures on their implementation.