Search

It is crucial for doctors to fully understand the interaction between drugs in prescriptions, especially when a patient takes multiple medications at the same time during treatment. The purpose of drug drug interaction (DDI) extraction is to automatically obtain the interaction between drugs from biomedical literature. Current state-of-the-art approaches for DDI extraction task are based on artificial intelligence and natural language processing. While such existing DDI extraction methods can provide more knowledge and enhance the performance through external resources such as biomedical databases or ontologies, due to the difficulty of updating, these external resources are delayed. In fact, user generated content (UGC) is another kind of external medical resources that can be quickly updated. We are trying to use UGC resources to provide more available information for our deep learning DDI extraction method. In this paper, we present a DDI extraction approach through a new attention mechanism called full-attention which can combine the UGC information with contextual information. We conducted a series of experiments on the DDI 2013 Evaluation dataset to evaluate our method. Experiments show improved performance compared with the state of the art and UGC-DDI model achieves a competitive F-score of 0.712. [ABSTRACT FROM AUTHOR]