Classifying medical relations in clinical text via convolutional neural networks
This work addresses relation extraction for clinical records, which is an incremental improvement over prior methods.
The study tackled medical relation classification in clinical text by proposing a CNN with multi-pooling and a constrained loss function, achieving competitive performance with existing ensemble methods on the 2010 i2b2/VA corpus.
Deep learning research on relation classification has achieved solid performance in the general domain. This study proposes a convolutional neural network (CNN) architecture with a multi-pooling operation for medical relation classification on clinical records and explores a loss function with a category-level constraint matrix. Experiments using the 2010 i2b2/VA relation corpus demonstrate these models, which do not depend on any external features, outperform previous single-model methods and our best model is competitive with the existing ensemble-based method.