58,395 articles related to executive functions and 447,123 articles related to intelligence

Data for this study were obtained using the Web of Science (WoS) Core Collection database. The two data pools 'executive function' and 'intelligence' were set up as the respective main search terms. The literature was limited to academic articles by which the analysis included only those listed in SSCI or SCI(E). No publication year was limited. Based on these criteria 58,395 articles related to executive functions and 447,123 articles related to intelligence were collected. This data collection was conducted on May 1, 2025. From the metadata of the literature, the analysis included the title, the abstract, and the author keywords. At the pre-processing stage, NLP (Natural Language Processing) technology was employed to increase the efficiency and consistency of the analysis. Indeed, all text was lower case, special characters and stopwords were deleted, and the text was tokenized. Additionally, the semantic network-based embedding function was used to normalize meaning-based keywords and embed synonyms to maintain semantic consistency among concepts and maximize the reliability of interpretation of results.

本研究的数据源自Web of Science（WoS）核心合集数据库。本次研究构建“执行功能”与“智力”两个主题词池，分别作为核心检索词。分析仅纳入SSCI或SCI(E)收录的学术论文，且未设置出版年份限制。按上述标准共采集到与执行功能相关的文献58395篇，与智力相关的文献447123篇，数据采集工作于2025年5月1日完成。从文献元数据中，本次分析提取了标题、摘要与作者关键词作为研究字段。在预处理阶段，采用自然语言处理（Natural Language Processing，NLP）技术以提升分析效率与一致性：具体包括将所有文本转换为小写格式，移除特殊字符与停用词，并完成Token化处理。此外，还采用基于语义网络的嵌入函数对语义类关键词进行规范化处理，并对同义词进行嵌入，以维持概念间的语义一致性，最大化研究结果解读的可靠性。