Clarivate report underlines need for structured data in research intelligence
Reliable research intelligence depends on more than scale or openness alone, according to a new report from Clarivate and the Institute for Scientific Information (ISI). As the volume and complexity of the global scholarly record continues to grow, the report warns that poorly structured or inconsistent data can undermine benchmarking, evaluation and policy decisions.
“For research to be discovered, compared and evaluated effectively, the underlying data must be clean, consistently structured and stable over time,” the report states. Without this foundation, gaps in data quality can lead to “distorted collaboration indicators,” “misleading comparisons” and “unstable trends.”
The report examines how deliberate data curation, stable categorisation and rich metadata support the Web of Science and its role in research discovery and analysis. “Good research analysis… depends on validated and structured data,” it says, adding that comprehensive metadata enhances interoperability across complementary data sources.
The origins of this approach date back to 1956 with the launch of Current Contents, which introduced structured indexing of journal literature using key metadata such as author, location and keywords. As research outputs and formats have diversified, ISI and Clarivate have adapted their systems to account for disciplinary differences in publishing practices and research culture.
“To make comparisons like-for-like, such differences need to be accounted for,” the report explains, highlighting variation in the importance of journals, conference proceedings and books across subject areas.
The report also explores the evolution of classification systems, balancing stable, top-down categories with more dynamic, bottom-up approaches that group research by shared characteristics and citation patterns. It points to ongoing work with the Centre for Science and Technology Studies (CWTS) at Leiden University to develop more granular, topic-based structures.
Beyond discovery, the report considers how structured data can be aligned with national research assessment frameworks and global priorities such as the UN Sustainable Development Goals, helping connect research outputs to societal impact.
Ultimately, the report argues that structured research activity data is essential for credible research intelligence. “Structured, verified, and interoperable research activity data are not a convenience – they are a prerequisite for credible search, discovery and evaluation.”