Single versus dual document-type assignment: an empirical comparison of classification policies in Scopus and Web of Science databases
摘要
Assigning standardised labels–such as article, review, or conference paper–to scholarly publications, known as document-type (DT) classification, plays a fundamental role in information retrieval, bibliometric analysis, and research evaluation. However, policies governing this process differ significantly across major databases. Web of Science (WoS) allows documents to be assigned multiple DTs simultaneously, whereas Scopus restricts each document to a single DT designation. This study investigates the implications of these divergent approaches, focusing on WoS documents carrying dual-DT status, a scenario absent in Scopus. Through a manual examination of over one thousand dual-DT WoS documents, three recurring patterns were identified: (i) combinations capturing both content and container descriptors (e.g., book chapters, proceedings papers), (ii) temporary dual assignments linked to early-access status, and (iii) the handling of specialised DTs (e.g., data papers, retracted papers). Results show that dual-DT documents are more prone to classification errors, reflecting inherent complexity, and that similar issues occur in Scopus despite its single-DT constraint. While WoS’s multi-DT policy provides richer descriptive detail, it is also susceptible to inconsistencies. In contrast, Scopus’s single-DT approach simplifies classification but may lead to increased misclassification, particularly for documents originating from conference proceedings or journal special issues. Even though the corpus is relatively small, this study provides a framework for expanding comparative analyses of different database classification policies. Clearer DT definitions, combined with flexible multi-DT classification options, could improve both accuracy and transparency in bibliometric indexing.