In order to avoid the ambiguous classification of articles in multiple categories in the Web of Science and the resulting complication of bibliometric indicators, a reclassification of articles in the Web of Sciences categories was carried out according to the method of S. Milojevi\'c (2020). The higher hierarchical level from the OST classification into 11 scientific disciplines is also revised. Though in most cases articles are assigned to a subject category close to the original category, the reclassification changes the subject category of about 50% of the documents of the database. Therefore, the world distribution of disciplines and disciplinary profiles of scientific actors are modified. A sample of twenty five countries highlights the impact of the reclassification on country specialization indexes. Field-normalized indicators are also impacted. The level of changes is explored in the case of the Mean Normalized Citation Indicator (MNCS). A more in-depth analysis of the MNCS in Mathematics is carried out and reveals different strategies of countries to publish works with a mathematical background.

翻译：为避免Web of Science中文章多类别归属导致的分类模糊性及其对文献计量指标造成的复杂性，本研究依据S. Milojević（2020）的方法对Web of Science学科类别进行了文章重分类。同时基于OST分类体系将更高层级的11个科学学科门类进行了修订。尽管多数文章被分配至与原类别相近的学科类别，此次重分类仍改变了数据库中约50%文献的学科归属。因此，全球学科分布格局及科研主体的学科特征均发生改变。对二十五个国家的样本分析突显了重分类对国家专业化指数的显著影响。领域归一化指标同样受到影响。本研究以平均归一化引用指数（MNCS）为例探究了指标变动的程度，并通过对数学领域MNCS的深入分析，揭示了各国在发表数学背景研究成果方面的差异化策略。