نوع مقاله : مقاله پژوهشی
نویسندگان
1 گروه علم اطلاعات و دانششناسی، واحد تهران شمال، دانشگاه آزاد اسلامی، تهران، ایران
2 استادیار گروه مهندسی کامپیوتر، دانشگاه شهاب دانش، قم، ایران
3 گروه بازیابی اطلاعات و دانش، واحد اراک، دانشگاه آزاد اسلامی، اراک، ایران
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
Purpose: The advent of Artificial Intelligence (AI) has fundamentally transformed Information Retrieval (IR) systems. Scientific publication databases, the vital arteries of the global research ecosystem, lie at the core of this technological revolution. Today, AI-driven capabilities—such as semantic search, user personalization, and intelligent analytics—have elevated researcher expectations, transitioning from digital novelties to indispensable standards for leading platforms. However, the global velocity of adopting these technologies remains asymmetrical. Evidence indicates a profound technological divide between international scientific databases and national counterparts. This disparity is not merely a technical deficiency; it represents a strategic challenge. Left unaddressed, this gap threatens to undermine domestic scientific authority, increase reliance on foreign platforms, and precipitate a massive outflow of national research data. Recognizing this, the primary objective of this study is to systematically identify, statistically analyze, and elucidate the dimensions of this AI adoption gap between domestic and international scientific journal databases, relying on a rigorous expert-driven assessment.
Method: Methodologically, this applied research employs an analytical-survey design framed within a gap analysis approach. The primary data collection instrument was a structured questionnaire encompassing 33 previously validated AI components relevant to IR systems. Empirical evaluation targeted prominent databases. The domestic group comprised major Iranian platforms: Irandoc, ISC, and Magiran. The international reference group included globally renowned platforms: ScienceDirect, Elsevier, and Emerald. Operational evaluation was conducted through hands-on assessments by a purposively selected panel of 19 experts in information science, AI, and scientometrics. Inferential statistical analyses were executed using the non-parametric Mann-Whitney U test to compare the performance of the independent database groups, supplemented by the effect size index (r) to determine the significance of the observed technological discrepancies.
Findings: Data analysis definitely confirmed a pervasive and statistically significant technological gap between the two categories. Out of the 33 theoretically effective AI components initially identified, evaluation revealed only 16 components had been operationalized into the assessed user interfaces. Consequently, comparative analysis strictly focused on these 16 active features. Findings demonstrated that in 11 components (approximately 69%), international databases exhibited statistically significant superiority over domestic counterparts. This technological chasm was particularly acute in functional areas impacting research quality. Supremacy of foreign platforms was unequivocally evident in critical components like "spell checking and error correction" (p = 0.000), "automated summarization and information extraction" (p = 0.000), "individual recommender systems" (p = 0.006), and "intelligent suggestion engines" (p = 0.007). A deep divide also existed in components vital for safeguarding academic integrity, including "plagiarism and duplication detection" (p = 0.001), "article quality assessment" (p = 0.009), and "citation network analysis" (p = 0.003). Conversely, statistical analysis revealed no significant difference in 5 specific AI components. These comprised "advanced intelligent search" (p = 0.443), "automated keyword extraction" (p = 0.068), and "analysis of authors and organizations" (p = 0.688). Parity in these domains suggests domestic databases have successfully implemented foundational search capabilities, reaching an acceptable baseline standard.
Conclusion: In conclusion, the documented disparity transcends technological specifications; it represents a qualitative shift in service provision philosophy. International databases have evolved from passive "information repositories" into dynamic "intelligent research assistants," autonomously providing services like predictive trend analysis and automated validation. In stark contrast, domestic databases remain predominantly tethered to conventional repository functions. This stagnation poses a severe strategic threat. To bridge this divide, this study proposes a national roadmap in three phases. Phase One (Fundamental) concentrates on implementing baseline AI components, such as word disambiguation and error correction. Phase Two (Developmental) focuses on integrating value-creating mechanisms, particularly intelligent recommender systems and citation analytics. Phase Three (Strategic) aggressively targets long-term investments in forward-looking capabilities, like predicting emerging scientific trends. Ultimately, integrating AI into domestic scientific databases is an existential necessity to safeguard national scientific sovereignty and solidify sustainable development foundations.
کلیدواژهها [English]
)
ارسال نظر در مورد این مقاله