Insights into Analogy Completion from the Biomedical Domain

Denis Newman-Griffis, Albert M. Lai, Eric Fosler-Lussier

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

6 Scopus citations

Abstract

Analogy completion has been a popular task in recent years for evaluating the semantic properties of word embeddings, but the standard methodology makes a number of assumptions about analogies that do not always hold, either in recent benchmark datasets or when expanding into other domains. Through an analysis of analogies in the biomedical domain, we identify three assumptions: that of a Single Answer for any given analogy, that the pairs involved describe the Same Relationship, and that each pair is Informative with respect to the other. We propose modifying the standard methodology to relax these assumptions by allowing for multiple correct answers, reporting MAP and MRR in addition to accuracy, and using multiple example pairs. We further present BMASS, a novel dataset for evaluating linguistic regularities in biomedical embeddings, and demonstrate that the relationships described in the dataset pose significant semantic challenges to current word embedding methods.

Original languageEnglish
Title of host publicationBioNLP 2017 - SIGBioMed Workshop on Biomedical Natural Language Processing, Proceedings of the 16th BioNLP Workshop
PublisherAssociation for Computational Linguistics (ACL)
Pages19-28
Number of pages10
ISBN (Electronic)9781945626593
StatePublished - 2017
Event16th SIGBioMed Workshop on Biomedical Natural Language Processing, BioNLP 2017 - Vancouver, Canada
Duration: Aug 4 2017 → …

Publication series

NameBioNLP 2017 - SIGBioMed Workshop on Biomedical Natural Language Processing, Proceedings of the 16th BioNLP Workshop

Conference

Conference16th SIGBioMed Workshop on Biomedical Natural Language Processing, BioNLP 2017
Country/TerritoryCanada
CityVancouver
Period08/4/17 → …

Fingerprint

Dive into the research topics of 'Insights into Analogy Completion from the Biomedical Domain'. Together they form a unique fingerprint.

Cite this