نوع مقاله : مقاله پژوهشی
نویسنده
گروه علم اطلاعات و دانش شناسی دانشگاه قم، دانشکده ادبیات و علوم انسانی، دانشگاه قم، قم، ایران
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسنده [English]
Objective: The primary objective of this research is to develop a novel, AI-driven framework for evaluating the quality of scientific databases. By benchmarking Persian databases against reputable international counterparts, the study maps the current landscape, proposes evidence-based strategies for service enhancement, and provides a reliable reference for researchers, database managers, and scientometric policymakers.
Methodology: This applied, descriptive-analytical study investigates active Persian and international scholarly indexing databases. The selected sample comprises Persian platforms—specifically the Scientific Information Database (SID), Magiran, and Noormags—and international counterparts, namely Web of Science and Scopus. Data were gathered by extracting bibliographic records and article metadata. Quality assessment criteria included data accuracy, comprehensiveness, coherence, metadata standardization, currency, retrievability, and duplicate record rates. AI and machine learning algorithms were employed for objective evaluation, specifically to detect metadata errors, identify duplicates, assess descriptive consistency, and evaluate retrieval performance.
Findings: The results indicate that international databases outperform Persian ones across most quality indicators. Data accuracy averaged 96% for Scopus and 95% for Web of Science, compared to an average of 85% for Persian databases. Metadata consistency scored 94% internationally versus 80% domestically, while duplicate record rates averaged 5% in Persian platforms. Furthermore, update latency was under three months for international databases, compared to approximately five months for Persian equivalents.
Conclusion: Findings demonstrate that AI-driven algorithms provide an effective, scalable approach for objective database quality assessment. While Persian databases play a vital role in disseminating indigenous scholarship, they require structural metadata upgrades, error and duplication reduction, accelerated indexing cycles, and intelligent automation to meet international standards. The proposed framework establishes a solid scientific foundation for database selection, policy refinement, and quality-of-service optimization in regional indexing systems
کلیدواژهها [English]
)
ارسال نظر در مورد این مقاله