What is the Research Topics Schema?
Research Topics are a classification system that is applied to content used in Web of Science Research Intelligence, such as publications, grants and patents.
The Research Topics Schema is based on the hierarchical structure of the Citation Topics schema used in InCites Benchmarking & Analytics. However, while the taxonomy is derived from the Citation Topics schema, the methodology used to assign topics is different (see the Methodology section below for more details).
The schema is a three-level hierarchical system:
- Macro level - The top level, contains 10 topics.
- Meso level - The 2nd level, contains 326 topics.
- Micro level - The 3rd level , contains 2478 topics.
You can see the full topic list here - Download the 2025 schema.
Methodology
While the Research Topics Schema uses the hierarchical structure and topic definitions from the Citation Topics Schema, the underlying methodology for assigning topics differs fundamentally.
Citation Topics are based on a clustering algorithm developed by CWTS Leiden, where clusters are identified using citation link analysis. Each cluster is then labeled using either manual or AI-assisted techniques, and a topic description is generated based on the article titles in a citation topic cluster (read more about it here. )
In contrast, Research Topics are assigned using semantic similarity analysis. A document (e.g., a publication, grant or patent) is analyzed for its textual content (such as title, abstract and keywords), and its similarity to each micro-level topic description in the taxonomy is calculated. Based on these similarity scores, the most relevant Research Topics are assigned.
How is it used?
Content used in the application is tagged with the 10 most closely related Micro topics from the schema, sorted in descending order of relevance, as identified by data science algorithms. The topics may or may not be from the same Meso topic.
When searching or filtering by Research Topics, only content that is tagged with this topic will be displayed or considered.
Researchers are tagged with 10 Micro-level topics that are aggregated from content related to the researchers, sorted in descending order of relevance.
When viewing researcher search results, the 3 most relevant Micro-level topics are displayed in descending order. When viewing a Research Card all 10 topics are displayed in descending order.
Where is it used?
Expert Finder
- Search by Research Topics - You can search for a researcher by selecting one or more Research Topics from any level.
- Researcher search results - When searching for researchers the search results include three topics per researcher.
- Researcher card - The researcher card displays 10 micro-level topics.
- Research Topics Filter - When searching for researchers by name or by an organization, there is an option to filter the results using one or more Research Topics.
For more information on using the Expert Finder, see here.
Research Performance Summary Report
- Filter by research topics - The Research Topics filter allows you to refine the set of publications and awarded grants included in the Research Performance Summary report. When applied, it limits the analysis to only those items associated with the selected research topics.
Frequently Asked Questions
How do Research Topics differ from Citation Topics?
Citation Topics are assigned based on citation network clustering, with each publication linked to exactly one topic. Research Topics, by contrast, are assigned using semantic similarity to topic descriptions and may result in multiple topic assignments per document. Additionally, Citation Topics are applicable only to publications, whereas Research Topics can be applied more broadly to other content types, including patents and grants.
Will a publication always include the Research Topic corresponding to its Citation Topic?
Yes, if a publication has a specific Citation Topic, the corresponding Research Topic is also assigned to ensure consistency across systems.
Can the Research Topics Schema evolve over time?
Yes, because topic assignment is based on semantic analysis rather than fixed citation clusters, the schema can be adapted over time. This includes refining the hierarchy, updating topic descriptions, or adding topics to better support research intelligence questions.
How is the Research Topics Schema applied in reporting?
Currently, the Research Topics Schema is used in reporting to filter publications. It is not yet used for normalized indicators. For example, the Category Normalized Citation Impact (CNCI) available in the organization summary report uses the Web of Science Categories to define subject fields. Research Topics are not yet used to group similar publications for normalization purposes. However, support for using Research Topics in normalization is planned for future development.
How is the Research Topics Schema used in Research Intelligence Assistant?
At present, Research Topics are not yet supported in Research Intelligence Assistant, neither for filtering content nor for indicator normalization. Integration is planned for a future release.