Folksonomy-Based Ontology: A Link Between Intelligence and Meaning in Psychology and Educational Sciences

• Published in: علوم و فنون مدیریت اطلاعات Vol. 10; no. 1; pp. 175 - 208
• Main Authors: Nosrat Goodarzi, Yaghoub Norouzi, Hosseini Beheshti Molouk Sadat, Mehdi Alipour-Hafezi, Behrooz Bayat
• Format: Journal Article
• Language: Persian
• Published: University of Qom 01-03-2024
• Subjects: educational sciences; folksonomy; information organization; ontology; psychology; semantic network; semi-automatic learning
• ISSN: 2476-6658; 2476-6534
• DOI: 10.22091/stim.2022.7865.1734

Purpose: The aim of the present study is to reconcile the dual embodiment of ontology and folksonomy and establish a symbiotic relationship between the two technologies for the development of folksonomized ontology as a tool for organizing information in the fields of psychology and educational sciences in cyberspace.Method: This study is of an applied nature and utilizes a combined research method. To gain conceptual knowledge and develop a semi-automated ontology based on natural language (common labels), the three-step method proposed by Alves and Santanche (extraction, enrichment, and synthesis) (2013) is utilized, along with Noy and McGuinness's approach (2001). The statistical population refers to the labels used by students of psychology and educational sciences in the digital library of Allameh Tabatabai University. The purposive sampling method is used to select and aggregate specific labels. Additionally, to evaluate the credibility and validity of the extracted concepts and relationships, they are reviewed and confirmed by 15 experts in the fields of psychology and educational sciences. The software used to create the ontology is Protégé version 5.5.Findings: A total of 15,787 labels were extracted. After identifying the fundamental concepts and their interrelationships using the Eric database, a conceptual framework with an ontological approach was developed. After confirming and validating the conceptual structure, an ontology based on folksonomy is presented as the main research finding. It consists of 277 classes and subclasses, 110 relation types, 50 samples, and 9 synonyms taken from the Eric database, and is illustrated in several graphs. Based on the statistical test performed on the ontology derived from folksonomy with a significance level of 0.0005, the model has been proven to be relatively accurate, and all components have been confirmed by experts.Conclusion: Folksonomies are "user-generated ontologies" within the Semantic Web. Leveraging the structural distinctions between folksonomies and ontologies holds the potential to advance the Semantic Web without requiring a deep understanding of complex classification systems.