Mobile Medical Question and Answer System with Improved Char-level based Convolution Neural Network and Sparse Auto Encoder
Guokai Yan, Jianqiang Li · 2019
Last few years, China has entered into an open and comprehensive two-child era which cause the demand for medical resources of elderly pregnant women more scarce. And the accuracy will be effected by the accuracy of Chinese word segmentation during the classification of questions. To solve those problem above, we are presenting a Chinese Mother-to-Child Domain Question Answering System. We propose a improved end-to-end convolutional neural networks with Sparse Auto Encoder(SAE) layer to extract information at Char-level. It has achieved a significant effect both in accuracy and avoiding over-fitting.