Reengineering of MeSH Thesauri for Term Selection to Optimize Literature Retrieval and Knowledge Reconstruction in Support of Stem Cell Research

Authors

Yan Su, Chinese Academy of Sciences
James E. Andrews, University of South FloridaFollow
hong huang, University of South FloridaFollow
Yue Wang, Chinese Academy of Sciences
Liangliang Kong, Chinese Academy of Sciences
Peter Cannon, University of South FloridaFollow
Ping Xu, Chinese Academy of Sciences

Document Type

Article

Publication Date

5-23-2016

Keywords

Knowledge organization system (KOS), MeSH, Thesaurus, Classify, Stem cells

Digital Object Identifier (DOI)

https://doi.org/10.1186/s12911-016-0298-z

Abstract

BACKGROUND: PubMed is a widely used database for scientists to find biomedical-related literature. Due to the complexity of the selected research subject and its interdisciplinary nature, as well as the exponential growth in the number of disparate pieces of biomedical literature, it is an overwhelming challenge for scientists to define the right search strategies and quickly locate all related information. Specialized subsets and groupings of controlled vocabularies, such as Medical Subject Headings (MeSH), can enhance information retrieval in specialized domains, such as stem cell research. There is a need to develop effective search strategies and convenient solutions for knowledge organization in stem cell research. The understanding of the interrelationships between these MeSH terms also facilitates the building of knowledge organization systems in related subject fields.

METHODS: This study collected empirical data for MeSH-related terms from stem cell literature and developed a novel approach that uses both automation and expert-selection to create a set of terms that supports enhanced retrieval. The selected MeSH terms were reconstructed into a classified thesaurus that can guide researchers towards a successful search and knowledge organization of stem cell literature.

RESULTS: First, 4253 MeSH terms were harvested from a sample of 5527 stem cell related research papers from the PubMed database. Next, unrelated terms were filtered out based on term frequency and specificity. Precision and recall measures were used to help identify additional valuable terms, which were mostly non-MeSH terms. The study identified 15 terms that specifically referred to stem cell research for information retrieval, which would yield a higher precision (97.7 %) and recall (94.4 %) rates in comparison to other approaches. In addition, 128 root MeSH terms were selected to conduct knowledge organization of stem cell research in categories of anatomy, disease, and others.

CONCLUSIONS: This study presented a novel strategy and procedure to reengineer term selections of the MeSH thesaurus for literature retrieval and knowledge organization using stem cell research as a case. It could help scientists to select their own search terms and build up a thesaurus-based knowledge organization system in interested and interdisciplinary research subject areas.

Rights Information

This work is licensed under a Creative Commons Attribution 4.0 License.

Citation / Publisher Attribution

BMC Medical Informatics and Decision Making, v. 16, art. 54

Scholar Commons Citation

Su, Yan; Andrews, James E.; huang, hong; Wang, Yue; Kong, Liangliang; Cannon, Peter; and Xu, Ping, "Reengineering of MeSH Thesauri for Term Selection to Optimize Literature Retrieval and Knowledge Reconstruction in Support of Stem Cell Research" (2016). School of Information Faculty Publications. 339.
https://digitalcommons.usf.edu/si_facpub/339