TY - JOUR
T1 - Automated disambiguation of acronyms and abbreviations in clinical texts
T2 - window and training size considerations.
AU - Moon, Sungrim
AU - Pakhomov, Serguei
AU - Melton, Genevieve B.
PY - 2012
Y1 - 2012
N2 - Acronyms and abbreviations within electronic clinical texts are widespread and often associated with multiple senses. Automated acronym sense disambiguation (WSD), a task of assigning the context-appropriate sense to ambiguous clinical acronyms and abbreviations, represents an active problem for medical natural language processing (NLP) systems. In this paper, fifty clinical acronyms and abbreviations with 500 samples each were studied using supervised machine-learning techniques (Support Vector Machines (SVM), Naïve Bayes (NB), and Decision Trees (DT)) to optimize the window size and orientation and determine the minimum training sample size needed for optimal performance. Our analysis of window size and orientation showed best performance using a larger left-sided and smaller right-sided window. To achieve an accuracy of over 90%, the minimum required training sample size was approximately 125 samples for SVM classifiers with inverted cross-validation. These findings support future work in clinical acronym and abbreviation WSD and require validation with other clinical texts.
AB - Acronyms and abbreviations within electronic clinical texts are widespread and often associated with multiple senses. Automated acronym sense disambiguation (WSD), a task of assigning the context-appropriate sense to ambiguous clinical acronyms and abbreviations, represents an active problem for medical natural language processing (NLP) systems. In this paper, fifty clinical acronyms and abbreviations with 500 samples each were studied using supervised machine-learning techniques (Support Vector Machines (SVM), Naïve Bayes (NB), and Decision Trees (DT)) to optimize the window size and orientation and determine the minimum training sample size needed for optimal performance. Our analysis of window size and orientation showed best performance using a larger left-sided and smaller right-sided window. To achieve an accuracy of over 90%, the minimum required training sample size was approximately 125 samples for SVM classifiers with inverted cross-validation. These findings support future work in clinical acronym and abbreviation WSD and require validation with other clinical texts.
UR - http://www.scopus.com/inward/record.url?scp=84880809454&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84880809454&partnerID=8YFLogxK
M3 - Article
C2 - 23304410
AN - SCOPUS:84880809454
SN - 0022-1120
VL - 2012
SP - 1310
EP - 1319
JO - Unknown Journal
JF - Unknown Journal
ER -