تدوین و توسعه هستی شناسی ارقام زراعی و باغی کشاورزی ایران با تأکید بر بازیابی ویدئو (مقاله علمی وزارت علوم)
درجه علمی: نشریه علمی (وزارت علوم)
آرشیو
چکیده
دانش بومی کشاورزی ایران وابسته به تنوع زیستی گیاهان مختلف زراعی و باغی است که در نقاط مختلف کشور، سازگار با آب و هوا و خاک منطقه و مقاوم به آفات، بیماری ها و علف های هرز است. هدف: به منظور معرفی گونه ها و ارقام مختلف که توسط پژوهشگران کشاورزی شناسایی شده و فراهم کردن امکان جستجو و بازیابی ویژگی های مختلف این ارقام و محل مناسب کاشت آنها برای کشاورزان هستان نگار IranAgriOnt طراحی گردیده است و به دلیل ماهیت دانش بومی کشاورزی در طراحی آن به بازیابی ویدئو توجه شده است. روش: به منظور توصیف ویژگی های بصری ویدئو از استاندارد MPEG7 استفاده شده است. الگوی هستان نگار ایران اگری اونت در نرم افزار پرتژه 5.5 و متشکل از سه بخش اصلی کلاس ها، ویژگی ها و نمونه-ها طراحی شده است.یافته ها: به طور کلی 26 کلاس در این هستان نگار ایجاد شده است که 20 مورد از آنها کلاس فرعی است. همچنین شامل 22 ویژگی شی و 33 ویژگی داده و شامل بیش از دویست نمونه است.Design of Ontology of Indigenous Knowledge on Agriculture in Iran with an emphasis on video retrieval
Introduction: Indigenous knowledge on agriculture in Iran depends on the biodiversity of various agricultural and garden plants, and the climate of each region. To introduce different indigenous species and cultivars that have been identified by agricultural researchers and provide the possibility of searching and retrieving various properties of these cultivars and the suitable cultivation area in Iran, IranAgriOnt was designed and due to the nature of indigenous knowledge on agriculture, for the design, video retrieval was considered. The present study has created the ontology of cultivars of agricultural and horticultural plants of Iran with the aim of using ontology in organizing the Indigenous Knowledge on Agriculture in Iran. In the design of this ontology for the purpose of vide retrieval, its related properties and relationships have been created. Among the extensive of indigenous agriculture knowledge due to the importance of biodiversity of plant species, This study identified the concepts and properties related to agricultural and horticultural cultivars in different regions of Iran. Methods: The ontology was designed using the content analysis. The ontology was assessed using qualitative assessment based on the criteria determined by the experts. To describe the visual properties of the video, MPEG7 has been used. IranAgriOnt is designed by Protege 5.5 and consists of three main parts: class, property and individual. For ontology design, the videos and written sources of indigenous knowledge on agriculture were searched and analyzed. next the concepts in the texts and videos and the relationships between them were identified. Sampling at this stage was purposive and continued until saturation was reached. The concepts, properties and relationships obtained were compared with the concepts found in the agricultural ontologies, thesaurus books and ontologies studied in the subject domain and video ontologies to select the correct concept and term. The classes, sub-classes and properties of objects and to some extent the properties of detected data related to video retrieval were adapted to MPEG7 and they were presented to the agricultural experts. After many trials and errors and holding numerous group or two-person meetings (subject expert, computer experts and researcher) of the classes and properties approved by the subject experts were arranged in order and the ontology obtained. Individuals were first entered the Excel and then the ontology. The computer experts tested the Iranian indigenous knowledge ontology by creating various queries in JavaScript. IranAgriOnt gave positive answers to various queries and was approved by the computer experts. The indigenous knowledge researchers evaluated the ontological model. At this stage, criteria-based approach was used and the agricultural experts evaluated and approved the designed ontology.Background: Related works were examined in two categories of Agricultural Indigenous Knowledge and video.Indigenous Knowledge and Agricultural ontologies were the Agricultural Ontology Service Concept Server (AOS/CS) designed by the Food and Agriculture Organization (FAO), the Crop ontology (CO) designed and used by the Consultative Group on International Agricultural Research (CGIAR) and The Plant Ontology (PO) created by the Plant Ontology Consortium and Agricultural gene ontology (AGO) is a subset of gene ontology (GO), which was created in order to recognize and analyze the genes of agricultural species to meet analysis demands from new technologies and research. Also, the countries of Spain, China, Nigeria, Sri Lanka, Malaysia, Thailand, Japan, Indonesia, India and Pakistan designed the ontology of indigenous agricultural knowledge. The works related to video ontology were also analyzed in order to identify the concepts and peroperties necessary in the design of video ontology from 2005 to 2018.Findings: In general, 26 classes have been created in this ontology, 20 of which are subclasses. It also includes 22 object properties, 33 data properties and more than two hundred individuals.In terms of using MPEG7, using the OWL and ontology design by Protégé, the present study is consistent with the related studies results. For the description of low- and high-level properties in the ontology, the study, has selected visual descriptors of color and shape. these properties are considered as properties of individuals of the class of Cultivar.In IranAgriOnt, did not find it appropriate to describe some low-level properties of the video, which were mentioned in MPEG7 for video description,
تبلیغات
