Poster Design & Printing by Genigraphics®- 800.790.4001Ted PedersenDepartment of Computer ScienceUniversity of Minnesota, ...
Upcoming SlideShare
Loading in …5
×

Pedersen naacl-2013-demo-poster-may25

422 views

Published on

Poster associated with UMLS::Similarity demo at NAACL 2013.

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
422
On SlideShare
0
From Embeds
0
Number of Embeds
7
Actions
Shares
0
Downloads
1
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Pedersen naacl-2013-demo-poster-may25

  1. 1. Poster Design & Printing by Genigraphics®- 800.790.4001Ted PedersenDepartment of Computer ScienceUniversity of Minnesota, Duluthtpederse@d.umn.eduhttp://www.d.umn.edu/~tpederseUMLS::Similarity is freelyavailable open sourcesoftware that allows a user tomeasure the semanticsimilarity or relatedness ofbiomedical terms found in theUnified Medical LanguageSystems (UMLS). It is writtenin Perl and can be used via acommand line interface, anAPI, or a Web interface.UMLS::Similarity has been modeled after andinspired by WordNet::Similarity (and yes, weveeven used some code). But, it has evolved to apoint where it is certainly more than a clone andhas its own very distinctive identity.The development of UMLS::Similarity wassupported in part by an RO1 grant from theNational Institutes of Health (USA), NationalLibrary of Medicine (#1R01LM009623-01A2).What are we measuring, and why?Similarity Depends on IS-A hierarchyAcknowledgmentsUsing UMLS::SimilarityAbstractContactUMLS::Similarity : Measuring the Relatedness and Similarity of Biomedical ConceptsBridget T. McInnes & Ying Liu : Minnesota Supercomputing InstituteTed Pedersen, Genevieve B. Melton & Serguei Pakhomov : University of Minnesotahttp://umls-similarity.sourceforge.netUnified Medical Language SystemTo be similar is to be alike, how much is Xlike Y? Similar concepts share ancestors inis-a hierarchy, the deeper the ancestor themore similar• LCS : least common subsumer●Tetanus and strep throat are similar, sinceboth are kinds of bacterial infectionsThe ability to organize concepts by theirsimilarity or relatedness to each other is afundamental operation in the human mind,and to many problems in Natural LanguageProcessing and Artificial IntelligenceUMLS::Similarity : Measuring the Relatedness and Similarity of Biomedical ConceptsBridget T. McInnes & Ying Liu : Minnesota Supercomputing InstituteTed Pedersen, Genevieve B. Melton & Serguei Pakhomov : University of Minnesotahttp://umls-similarity.sourceforge.netUMLS::Similarity : Measuring the Relatedness and Similarity of Biomedical ConceptsBridget T. McInnes & Ying Liu : Minnesota Supercomputing InstituteTed Pedersen, Genevieve B. Melton & Serguei Pakhomov : University of Minnesotahttp://umls-similarity.sourceforge.netUMLS::Similarity : Measuring the Relatedness and Similarity of Biomedical ConceptsBridget T. McInnes & Ying Liu : Minnesota Supercomputing InstituteTed Pedersen, Genevieve B. Melton & Serguei Pakhomov : University of Minnesotahttp://umls-similarity.sourceforge.netRelatedness Relies on DefinitionsAssign a numeric value that quantifies howsimilar or related two concepts or sensesare, not wordsCold may be temperature or illnessTo be related is much more general, sincethere are many ways to be related is-a,part-of, treats, symptom-of, ...●Tetanus and deep cuts are related butthey really arent similar (deep cuts cancause tetanus though)●Related words often defined using thesame ore similar words, look for overlapsWeb Interface• Allows for all measures to be computedusing a subset of possible sources•http://atlas.ahc.umn.edu•http://maraca.d.umn.eduCommand Line• Supports all measures, all UMLS sourcesplus many additional functions (many fromUMLS::Interface), examples include :•GetChildren•GetParents•GetRelated•GetSemanticGroup•FindCuiDepth•FindPathtoRoot•findLeastCommonSubsumerSemantic Similarity MeasuresPath basedShortest Path (path, cdist)Depth basedLeacock & Chodorow (lch)Zhong et al. (zhong)Nguyen & Al-Mubaid (nam)Information ContentResnik (res)Lin (lin)Jiang & Conrath (jcn)Relatedness MeasuresPath BasedHirst & St-Onge (hso)Definition BasedLesk (lesk)Adapted Lesk (lesk)Definition + CorpusGloss Vector (vector)The UMLS is a date warehouse distributed bythe National Library of Medicine (twice a year)It includes more than 100 terminologies, codesets, and ontologies encompassing manydifferent areas of medical knowledge. A usercan access individual sources (examplesbelow) or view them as one large combinedresource via the MetaThesaurus.MeSH – medical subject headings, used forindexing articles in PubMedFMA – Foundational Model of Anatomy, avery fine grained ontology of human anatomyOMIM – Online Mendelian Inheritance inMan, catalog of genes and gene disordersSNOMEDCT – Systematized Nomenclatureof Medicine – Clinical TermsWord Sense Disambiguationwith UMLS::SenseRelateWe can measure senses, or we can use themeasures to identify senses!http://search.cpan/org/dist/UMLS-SenseRelate

×