Send to

Choose Destination
See comment in PubMed Commons below
AMIA Annu Symp Proc. 2007 Oct 11:588-92.

Decompositional terminology translation using network analysis.

Author information

Department of Biomedical Informatics, Columbia University, New York, NY, USA.


Biomedical terminologies often contain composite concepts that cannot be translated into single unique synonymous concepts in a target controlled terminology. Such composite concepts need to be decomposed into sets of component concepts present in the target terminology that can serve as the proxy for applications in information retrieval, decision support or data analysis. Towards this goal, we use a "clustering coefficient" over the UMLS Metathesaurus to traverse the closely clustered neighbors of the composite source concept to generate a ranked list of possible component concepts. Using the MeSH Associated Expression mappings as the gold-standard, we show that the proposed approach generates relevant component concepts as compared to existing semantic locality based methods. The topological connectivity of the concepts in the UMLS Metathesaurus is a useful feature that can be coupled with existing lexical and semantic locality based approaches towards terminology translation.

[Indexed for MEDLINE]
Free PMC Article
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for PubMed Central
    Loading ...
    Support Center