Ellakwa, S., Elkafrawy, P. (2017). Establishing Dynamic Ontology for Agriculture Domain. The Egyptian Journal of Language Engineering, 4(2), 41-58. doi: 10.21608/ejle.2017.59443
Suzan F. Ellakwa; Passent M. Elkafrawy. "Establishing Dynamic Ontology for Agriculture Domain". The Egyptian Journal of Language Engineering, 4, 2, 2017, 41-58. doi: 10.21608/ejle.2017.59443
Ellakwa, S., Elkafrawy, P. (2017). 'Establishing Dynamic Ontology for Agriculture Domain', The Egyptian Journal of Language Engineering, 4(2), pp. 41-58. doi: 10.21608/ejle.2017.59443
Ellakwa, S., Elkafrawy, P. Establishing Dynamic Ontology for Agriculture Domain. The Egyptian Journal of Language Engineering, 2017; 4(2): 41-58. doi: 10.21608/ejle.2017.59443
Establishing Dynamic Ontology for Agriculture Domain
1Department in Central Laboratory for Agricultural Expert System (CLAES), Agricultural Research Center (ARC), Giza, Egypt
2Mathematics and Computer Science Department, Faculty of Science, Menoufia University
Abstract
This paper presents semi-automated system for establishing integrated ontology by merging two ontologies. It uses two processes: matching and merging. Matching process uses string-based technique, this technique uses four methods: exact method to detect identical terms, and substring, suffix and prefix methods to compare between terms. Using these four methods altogether improve the effectiveness of matching process, matching process uses also language-based techniques; this technique uses WordNet Method to detect terms that have the same meaning. This technique improves also the effectiveness of matching process. The proposed system presents a merging method of taxonomies in effective way. The system solves redundancy and inconsistency problem in integrated ontology.The proposed system is applied on the agricultural domain for Faba Bean crop to get an integrated ontology, it can be applied also on all crops whatever field crops or horticulture crops. The evaluation of the system shows that the performance of the system has high quality. The comparison of the proposed system and other systems shows that the proposed system has advantage of using five matching methods for mapping between terms that make the mapping between terms more perfect and efficient. The merger algorithm solves problems which appeared in other systems.
View on SCiNiTO
Top