Linking Medications and Their Attributes in Clinical Notes and Clinical Trial Announcements for Information Extraction: A Sequence Labeling Approach

Abstract

The goal of this work is to evaluate binary classification and sequence labeling methods for medication-attribute linkage detection in two clinical corpora. The results show that with parsimonious feature sets both the Support Vector Machine (SVM)-based binary classification and Conditional Random Field (CRF)-based multi-layered sequence labeling methods are achieving high performance.