【摘要】 随着科技的发展,智能医疗已经成为当下学界的热点研究内容。本文主要研究的是呼吸内科疾病的智能诊断,使用电子病历中的症状实体和异常检查结果实体来诊断患者可能患有的疾病。本文比较了不同的模型在该任务上表现,包括传统机器学习和深度学习。并且在深度模型中加入了不同的图表示学习方法以及提出了注意力机制来加强疾病和症状之间的联系。在实验中,本文提出的结合注意力机制和卷积神经网络以及外部向量获得了最优秀的表现。更多还原

【Abstract】 With the development of science and technology,intelligemt medical treatment has become a hot research topic in the current academic circles. This paper focuses on the intelligent diagnosis of respiratory diseases,using symptomatic entities and abnormal test results entities in electronic medical records to diagnose diseases that patients may have. The paper compares the performance of different models on this task, including traditional machine learning and deep learning. In addition, graph representation learning methods are added to the deep learning and attention,which is used to strengthen the relationship between disease and symptoms. In the experiment,the model which combines attention,Convolutional Neural Network(CNN) and external vector achieve the best performance.更多还原