Open AccessAN ONTOLOGY-BASED KNOWLEDGE REPRESENTATION USING ANALYTIC HIERARCHY PROCESS FOR ENHANCING SELECTION OF PRODUCT PREFERENCES
Author(s) -
O. R. VINCENT,
A. OWOADE,
A. S. MAKINDE,
A. T. AKINWALE
Publication year - 2017
Publication title -
journal of natural sciences, engineering and technology/journal of natural science, engineering and technology
Language(s) - English
Resource type - Journals
eISSN - 2315-7461
pISSN - 2277-0593
DOI - 10.51406/jnset.v16i1.1799
Subject(s) - analytic hierarchy process , selection (genetic algorithm) , judgement , computer science , ontology , product (mathematics) , fuzzy logic , consistency (knowledge bases) , data mining , order (exchange) , representation (politics) , artificial intelligence , information retrieval , machine learning , operations research , mathematics , economics , philosophy , geometry , epistemology , finance , politics , political science , law
Product alternatives, which emerges from large number of websites during searching, accounts for some hesitation experienced by customers in selecting satisfying product. As a result, making useful decision with many trade-off considerations becomes a major cause of such problem. Several approaches have been employed for product selection such as, fuzzy logic, Neuro-fuzzy, and weighted least square. However, these could not solve the problem of inconsistency and irrelevant judgement that occur in decision making. In this study, Ontology-based Analytic Hierarchy Process (AHP) was used for enhancing selection of product preferences. The model involved three fundamental components: product gathering, selection and decision making. Ontology Web Language (OWL) was utilized to define ontology in expressing product information gathering in a standard and structured manner for the purpose of interoperability while AHP was employed in making optimal choices. The procedure accepts customers’ perspectives as inputs which are classified into criteria and sub-criteria. Owl was created to foster customers’ interaction and priority estimation tool for AHP in order to generate the consistency ratio of individual judgements. The model was benchmarked with Geometric Mean (GM), Eigenvector (EV), Normalized Column Sum (NCS) Weighted Least Square (WLS) and Fuzzy Preference Programming (FPP). First and second order total deviations and violation rate were the performance parameters evaluation with AHP. The results showed that the minimum and maximum units of products are 2,452and 3,574, respectively. These implied that the proposed model was consistent, relevant and reflected a non-violation of judgment in selection of product preferences.