Abstract
BackgroundBiomedical document triage is the foundation of biomedical information extraction, which is important to precision medicine. Recently, some neural networks-based methods have been proposed to classify biomedical documents automatically. In the biomedical domain, documents are often very long and often contain very complicated sentences. However, the current methods still find it difficult to capture important features across sentences.ResultsIn this paper, we propose a hierarchical attention-based capsule model for biomedical document triage. The proposed model effectively employs hierarchical attention mechanism and capsule networks to capture valuable features across sentences and construct a final latent feature representation for a document. We evaluated our model on three public corpora.ConclusionsExperimental results showed that both hierarchical attention mechanism and capsule networks are helpful in biomedical document triage task. Our method proved itself highly competitive or superior compared with other state-of-the-art methods.
Highlights
Biomedical document triage is the foundation of biomedical information extraction, which is important to precision medicine
Our experimental result shows its effectiveness in improving the performance of the document triage task
Datasets and evaluation metrics In our experiment, we used the corpus from the BioCreative II Interaction Article Sub-task, BioCreative III Article Classification Tasks and the BioCreative VI Precision Medicine Track to mine for protein interactions and mutations for a precision medicine task
Summary
Biomedical document triage is the foundation of biomedical information extraction, which is important to precision medicine. Biomedical document triage is an important task in BioNLP, and is the first step in the literature curation workflow [4, 5]. Biomedical document triage helps curators and researchers focus on the biomedical literature that contains information relevant to their tasks [6, 7]. Biomedical document triage has been an important shared task in the BioCreative challenge community. Various methods have been proposed for the task of biomedical document triage [11]. The majority of these tasks can be divided into either machine learning-based methods or neural network-based methods
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
