Using Center Vector and Drug Molecular Information for Drug Drug Interaction Extraction
Biao Duan, Lei Qin, Jing Peng · 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · 2021
When a patient uses multiple drugs, the interaction between the drugs may cause changes in efficacy or produce unknown side effects, therefore, attention should be paid to the use of the characteristics of various drugs when combining drugs in clinical practice. The purpose of Drug Drug Interactions (DDI) is to extract the specific relationships of the drug pairs in the medical text. At present, the methods based on pre-training model has a good effect, but these methods still have room to improve, Here we proposed a novel method to extract DDIs, and conducted a series of experiments on the DDIExtraction 2013 dataset to evaluate our approach, experiments show that our method is superior to other most advanced methods, reaching the highest F1-score of 84.47 %.