259
Electronic Engineering and Information Science – Wang (Ed.)
© 2015 Taylor & Francis Group, London, ISBN: 978-1-138-02772-5
A method based on word feature selection for relation extraction in EMRs
X.B. Lv, Y. Guan & J.W. Wu
School of Computer Science and Technology, Harbin Institution of Technology, Harbin, Heilongjiang, China
ABSTRACT: Relations between entities in electronic medical records play an important role in understand-
ing the content of EMRs. A CRF-based machine learning method was applied in this paper. In order to select
a better feature set, we first merged the jargon words and common words in the preprocessing step, then dis-
criminate category matching and entropy-based feature ranking was used to optimize the features selected. ...