Mapping the research topics of NIH intramural research
Wednesday, September 16, 2015 — Poster Session I
- CW Belter
In this poster, we combined citation- and text-based methods to identify the research topics of papers published by intramural researchers at 10 NIH institutes. Our goal was to identify both the unique strengths of each institute and potential areas of topical overlap among institutes with minimal human intervention. We randomly selected 10 NIH institutes for analysis and retrieved the article metadata for all of their intramural research publications from 2008 to 2012 from the Web of Science article database. For each institute, we loaded the institute’s publication metadata into the Science of Science Tool (Sci2), created a bibliographic coupling network on their publications, and clustered the network using a community detection algorithm. We then performed word co-occurrence analysis on the titles of the papers assigned to each cluster to identify the topics of the papers in each cluster. If more than one major subject area was identified, we manually consolidated them into a broader concept to represent the research topic of the cluster. Preliminary results show that the 10 NIH institutes pursue research on both unique and shared research topics. Examples of cross-institute research topics include genetics, neuroscience, and cancer research. As we expand our analysis to include all of the major institutes at NIH, additional topics may emerge. Our results demonstrate the utility of our method and provide insights that could be used to facilitate communication and collaboration across the NIH intramural research community.
Category: Research Support Services