The swift growth in the form of digital information stored in the biomedical databases in this digital era has activated a prototype shift in the models in the Deep learning approaches which have used in several contests in Machine Learning and in the domain of pattern recognition. Finding relationship between entities like genes, diseases, proteins and drugs is tedious task due the ambiguity in the terms used in biomedical domain. Treating cancer with the drug based on the gene that is associated with the disease increases the survival rate. Hence, the deep learning method “Generative Adversarial Networks” is proposed to find the relationship between Genes, Diseases and Drugs from biomedical abstracts in this work.

在这个数字时代，存储在生物医学数据库中的数字信息形式的迅速增长激活了深度学习方法模型的原型转变，该方法已在机器学习和模式识别领域的多次竞赛中使用。由于在生物医学领域使用的术语含糊不清，因此在诸如基因，疾病，蛋白质和药物之类的实体之间寻找关系是一项繁琐的任务。使用基于与疾病相关的基因的药物治疗癌症可以提高生存率。因此，在这项工作中，提出了一种深度学习方法“生殖对抗网络”，以从生物医学摘要中找到基因，疾病和药物之间的关系。