AI在教育研究領域的應用系列（十七）：生成式AI輔助文獻探討與Scopus AI

Applications of AI in Educational Research Series (XVII): Generative AI-Assisted Literature Review and Scopus AI

隨著學術論文數量呈指數型增長，資訊過載問題日益嚴重。本研究旨在探討生成式人工智慧在學術文獻探索中的應用，並以Scopus AI為例，示範生成式人工智慧如何輔助研究者進行文獻綜整與知識探索。研究採用實際操作方式，以transformational leadership為檢索關鍵詞，針對Scopus AI（2025年12月版本）之核心功能進行系統性介紹與實測，包括自然語言查詢、主題摘要生成、概念視覺化、主題專家識別及新興主題分析等功能。本研究認為生成式AI宜定位為輔助性工具而非替代性工具，須在人類監督下使用此類系統，遵循透明揭露原則，並回歸原始文獻進行批判性檢視，以維持學術誠信與研究品質。

With the exponential growth of scholarly publications, information overload has become an increasingly pressing challenge. This study investigates the use of generative artificial intelligence for academic literature discovery and, using Scopus AI as an illustrative case, shows how generative AI can support researchers in synthesizing prior work and exploring disciplinary knowledge. Adopting a hands-on approach, we used “transformational leadership” as the search term and provided a systematic overview and empirical trial of the core features of Scopus AI (December 2025 release), including natural-language querying, topic-summary generation, concept visualization, topic-expert identification, and emergingtopic analysis. We contend that generative AI should be framed as a supportive rather than substitutive tool: it should be deployed with human oversight, accompanied by transparent disclosure, and grounded in critical verification against the original sources to safeguard academic integrity and research quality.