| چکیده انگلیسی مقاله |
Purpose: The purpose of this study was to identify the factors affecting the improvement of relevance in information retrieval, as perceived by faculty members of the Department of Information Science and Knowledge in the Humanities and Medical Librarianship departments, within the scientific social networks of LinkedIn and ResearchGate. Method: This research is applied in nature and was conducted using a survey methodology.
The study population sizes in the three departments—Humanities, Medical Librarianship, and Comprehensive Sciences—are 110, 102, and 207 individuals, respectively. Based on the Cochran formula, the sample sizes for statistical analysis are 86, 81, and 135 individuals, respectively, all of whom have utilized the scientific networks LinkedIn and ResearchGate. The selection of these two scientific networks was based on their higher number of referrals and the involvement of several experts. To assess the reliability of the questionnaire, Cronbach's alpha coefficient was employed. Additionally, the composite reliability method was utilized to evaluate the reliability of the constructs. To establish the validity of the questionnaire, input from various experts and specialists in management, library and information science, and sociology was considered. The extracted mean variance index was used to measure validity (credibility). The analysis of the structural model fit was conducted through confirmatory factor analysis and equation modeling, utilizing SmartPLS software. Findings: The results of the regression analysis indicate that among the research components, the presence of feedback, with a coefficient of 0.775, has the most significant impact on relevance and its enhancement in information retrieval within the LinkedIn database. Following the feedback variable, the variables of requests and questions, user characteristics, and database characteristics rank next, with coefficients of 0.515, 0.492, and 0.471, respectively. In the meantime, the information system does not demonstrate a significant effect. However, based on the results of the regression analysis, among the research components, the presence of feedback, with a coefficient of 0.812, has the most substantial impact on relevance and its enhancement in information retrieval within the ResearchGate database. Following the feedback variable, the variables of document characteristics, user characteristics, and requests and questions rank next, with coefficients of 0.726, 0.608, and 0.541, respectively. Overall, the information system and retrieval system do not exhibit a significant effect. Conclusion: Based on the comprehensive research model, the effective factors identified in scientific social networks, such as the information system and retrieval system, can be regarded as a holistic framework for enhancing the relevance of information retrieval in these networks. |