An Extraction Method for Drug-Drug Relationships based on External Knowledge

Abstract:

Drug-Drug Interaction refers to the inhibition or promotion between drugs. In order to solve the problem that the extraction results of the current model are quite different in different relationship categories, this paper proposes a relationship extraction model based on external knowledge. This method first processes the information in the external drug database, constructs a data set with drug description information on which the model is trained and then the optimal model is saved. Finally, the resulting optimal model is combined with the drug relationship extraction model to extract drug relationships, which can make better use of the existing knowledge in drug database, alleviate the problem of large differences in the extraction results of different relationship categories, and improve the extraction result. The experimental results on the DDIExtraction 2013 dataset show that the F1 value of the proposed approach in this paper is 2.47% higher than the current best approaches.