What are Emerging Topics?
An Emerging Topic is identified by two components: 1) a core of closely related publications, and 2) a frontier of recently published papers that cite a paper in the core. Core papers are highly cited papers published in the last 5 years. They are clustered together if they are frequently co-cited by the same publications. A period of rapid citation activity around the core publications signifies there is something happening in that realm of research. This use of the citation network means that an Emerging Topic provides an intuitive, powerful, and ever-evolving way to identify contemporary topics of the highest interest spanning the entire published literature.
To be considered an Emerging Topic, the following must be true:
- There must be a minimum of 3 and a maximum of 25 core publications
- The average publication year for core publications must not exceed 5 years
- The average publication year for co-citing publications must not exceed 2.5 years
It is possible for a publication to appear in multiple Emerging Topics, but it is rare.
Generative AI is used to assign a title that accurately represents the publications within an emerging topic cluster. The topics are updated monthly to capture the latest trends and research developments.
Core Publication
A core publication is a highly cited paper ranking in the top 1% by citations received in the same field, year of publication, and document type and published within the past 5 year. The Leiden algorithm is used to cluster publications into coherent clusters. Core publications are clustered together if frequently co-cited by the same publications.
Co-citing Publication
A co-citing publication is a paper that cites two or more of the core publications. The average publication year for co-citing papers must be 2.5 years or less.
Interdisciplinarity Index
The Interdisciplinarity Index measures how broadly an emerging research topic integrates ideas from different fields. This index is calculated by examining the variety of Web of Science categories assigned to the cited publications of both the core and co-citing publications. These cited publications represent the knowledge sources that the emerging topic builds upon. A higher diversity in the Web of Science categories of these cited publications indicates greater interdisciplinarity.
The interdisciplinarity score ranges from 0 to 1. A score of 0 indicates that the cited publications are all within the same Web of Science category, showing no interdisciplinarity. A score of 1 indicates that the cited publications are evenly distributed across different Web of Science categories, showing maximum interdisciplinarity.
The index is calculated using Normalized Shannon Entropy, formulated as follows:
I(topic) = −∑n(pilog(pi)) / log(n)
Where:
- pi is the proportion of cited publications in Web of Science category i
- n is the total number of Web of Science categories (currently 254)
- The denominator, log(n), normalizes the score to a range between 0 and 1
This method ensures the Interdisciplinarity Index reflects how well an Emerging Topic integrates diverse ideas from various fields by analyzing the cited knowledge sources of core and co-citing publications.
Mean Publication Year
The Mean Publication Year is the average year of publication for all publications (both core and co-citing) within an emerging topic. Only publications from the last 5 full years, plus the current year to date, can be included in an emerging topic.
Mean Co-Citing Publication Year
The Mean Co-Citing Publication year is the average publication year for all co-citing publications with an emerging topic. The average publication year for co-citing publications must not exceed 2.5 years.
Publications
Publications is the total number of core and co-citing publications conforming the Emerging Topic. Some can be both core and co-citing, so the sum of both metrics might not correspond to the total publications.
Emerging Topic Category
An Emerging Topic Category is a high-level grouping of related emerging topics, designed to provide a broader thematic organization of research trends. Each category contains multiple emerging topics that share conceptual similarities, enabling exploration of research trends at both granular (topic) and aggregated (category) levels.
The categories are formed using advanced clustering techniques applied to the sentence embeddings of emerging topic titles. In total, there are 25 Emerging Topic Categories, each representing a significant research domain. Categories are exclusive, meaning that an emerging topic can belong to only one category.
Unlike emerging topics, which are updated monthly, emerging topic categories are updated annually to reflect long-term shifts in research trends and themes.
Examples of Emerging Topic Categories include:
- Machine learning and AI applications
- Advanced materials engineering
- Environmental and climate impacts
- Global health and disease trends
- Nanotechnology in drug delivery
- Education, cognitive science, and technology
- Social and political dynamics
Refer to the complete list of categories in the Emerging Topics Research Area Schema documentation for further details.
Primary Category
Each Emerging Topic is assigned a primary category. The primary category comes from the 254 Web of Science subject categories. The assigned primary category is the Web of Science subject category that appears most frequently across all publications within an Emerging Topic.
Secondary Category
If at least 25% of publications within an Emerging Topic are assigned to a second Web of Science subject category, that category will appear as a Secondary Category.
What questions can be answered?
Emerging Topics helps you answer questions like:
- What are new and emerging areas of research?
- Which emerging topics show high interdisciplinarity?
- Who are the top organizations and authors involved in these emerging topics?
- Which funding agencies are supporting emerging areas?
What data is used by Emerging Topics?
Content in Emerging Topics comes from the last 5 year of Web of Science Core Collection. This includes SCIE, SSCI, AHCI, ESCI, and CPCI. It includes articles, reviews, and conference proceedings.
The addition of Conference Proceedings (CPCI) results in more topics in Computer Science & Engineering.