Word sense disambiguation across two domains: Biomedical literature and clinical notes

  • Authors:

  • Guergana K. Savova;Anni R. Coden;Igor L. Sominsky;Rie Johnson;Philip V. Ogren;Piet C. de Groen;Christopher G. Chute

  • Affiliations:

  • Division of Biomedical Informatics, Mayo Clinic College of Medicine, 150 Third Street SW, Rochester, MN 55902, USA;IBM, T.J. Watson Research Center, Hawthorne, New York, USA;IBM, T.J. Watson Research Center, Hawthorne, New York, USA;RJ Research Consulting, Tarrytown, New York, USA;Division of Biomedical Informatics, Mayo Clinic College of Medicine, 150 Third Street SW, Rochester, MN 55902, USA;Division of Biomedical Informatics, Mayo Clinic College of Medicine, 150 Third Street SW, Rochester, MN 55902, USA;Division of Biomedical Informatics, Mayo Clinic College of Medicine, 150 Third Street SW, Rochester, MN 55902, USA

  • Venue:
  • Journal of Biomedical Informatics
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

The aim of this study is to explore the word sense disambiguation (WSD) problem across two biomedical domains-biomedical literature and clinical notes. A supervised machine learning technique was used for the WSD task. One of the challenges addressed is the creation of a suitable clinical corpus with manual sense annotations. This corpus in conjunction with the WSD set from the National Library of Medicine provided the basis for the evaluation of our method across multiple domains and for the comparison of our results to published ones. Noteworthy is that only 20% of the most relevant ambiguous terms within a domain overlap between the two domains, having more senses associated with them in the clinical space than in the biomedical literature space. Experimentation with 28 different feature sets rendered a system achieving an average F-score of 0.82 on the clinical data and 0.86 on the biomedical literature.