AI 中文总结
研究比较Web of Science、Scopus、Dimensions和The Lens四个主要文献计量平台分类方案，分析其方法、结构和粒度级别差异，助研究人员依自身目标选合适平台，提升文献计量分析准确性与相关性。
AI 中文摘要
本研究考察了四个主要文献计量平台——Web of Science、Scopus、Dimensions和The Lens——的分类方案，这些平台共索引了数亿篇学术文献。随着这些平台规模和范围的扩大，研究人员在搜索和分析其大量文献时面临日益增加的复杂性。分类方案是浏览这些海量数据集的重要工具，能让研究人员按领域或主题高效筛选出版物。然而，各平台分类方法差异巨大，从专家策划的期刊级别分类到人工智能驱动的文献级别分类，这会显著影响研究结果。本研究全面比较了各平台的分类方法、结构和粒度级别。通过了解这些差异，研究人员能更明智地决定哪个平台的分类系统最符合其特定研究目标，最终提高文献计量分析的准确性和相关性。
英文摘要
This study examines the classification schemes of four major bibliometric platforms - Web of Science, Scopus, Dimensions, and The Lens - which collectively index hundreds of millions of scholarly documents. As these platforms grow in scale and scope, researchers face increasing complexity when searching and analyzing their vast collections. Classification schemes serve as essential tools to navigate these massive datasets, enabling researchers to efficiently filter publications by field or topic. However, the substantial differences between platforms' classification approaches - ranging from expert-curated journal-level categorization to AI-driven document-level classification - can significantly impact research outcomes. This study provides a comprehensive comparison of each platform's classification methods, structures, and granularity levels. By understanding these variations, researchers can make more informed decisions about which platform's classification system best aligns with their specific research objectives, ultimately improving the accuracy and relevance of their bibliometric analyses.
