Search PubMed⌕ Search
Home/Archive/PubMed

PubMed · 11825228

Evaluating UMLS strings for natural language processing.

A T McCray·O Bodenreider·J D Malley·A C Browne
View original source ↗Export metadata ↓

Abstract

The National Library of Medicine's Unified Medical Language System (UMLS) is a rich source of knowledge in the biomedical domain. The UMLS is used for research and development in a range of different applications, including natural language processing (NLP). In this paper we investigate the nature of the strings found in the UMLS Metathesaurus and evaluate them for their usefulness in NLP. We begin by identifying a number of properties that might allow us to predict the likelihood of a given string being found or not found in a corpus. We use a statistical model to test these predictors against our corpus, which is derived from the MEDLINE database. For one set of properties the model correctly predicted 77% of the strings that do not belong to the corpus, and 85% of the strings that do belong to the corpus. For another set of properties the model correctly predicted 96% of the strings that do not belong to the corpus and 29% of the strings that do belong to the corpus.

Explore related subjects

MEDLINEModels, StatisticalNatural Language ProcessingUnified Medical Language SystemVocabulary, ControlledJournal Article

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

A T McCray, O Bodenreider, J D Malley, A C Browne. 2001. Evaluating UMLS strings for natural language processing.. https://pubmed.ncbi.nlm.nih.gov/11825228/

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2006

The use of PubMed/Medline in psychiatry. 2: The PubMed search window: a description and a tutorial.

A previous paper, in a series of three, presented the National Library of Medicine, Medline, PubMed and the MeSH system. Connection to PubMed via the Internet opens the PubMed search window that is the tool for retrieving references to scientific medical literature. This paper is intended as a tutorial for an inexperienced visitor to PubMed, and describes the search window in some detail. Several proposals for exercises in its use are given in the text, while explanations and comments to the exercises are presented in an appendix. The following procedures are described: retrieving references, ways to see the abstract, ways of displaying, sorting and showing the references, together with how to produce lists, import references to your own computer, and ways to get the full text of a paper.

MEDLINE↗
PubMed2006

The use of PubMed/Medline in psychiatry. 3: Searching PubMed.

This paper is the third in a series of three, intended as a tutorial in the use of PubMed/Medline for an inexperienced user. The papers have the following contents: I--a description of NLM, Medline, PubMed and the system of Medical Subject Headings (MeSH), which form the basis for the indexing of scientific articles, books and other items at NLM. II--A description and a tutorial of the PubMed search window. III--The present article deals mainly with the searching for references in PubMed. Ways of restricting and concentrating the search are presented, and exercises are proposed. A reader may also find guidance for a search for medical books in the NLM Catalog, and in the use of tools like Related Articles, Bookshelf, and Index. With eating disorders as an example, more information is presented on the use of MeSH terms.

MEDLINE↗