基于神经网络的信息提取方法研究

Research on Information Extraction Method Based on Neural Network

神经网络现已应用于信息提取中,但在实际应用中存在组件间联系不密切、信息丢失严重等问题,因此需要优化神经网络模型。基于此,提出基于神经网络的信息提取方法,分别从人脸信息提取和动作信息提取两方面建立对应模型,并给出实验与分析结果。实验结果表明,该方法能够有效提取信息,具有较高的识别准确率。

Neural network has been applied in information extraction,but there are some problems in practical application,such as the relative relationship between components is not close,information loss is serious,so it is necessary to optimize the neural network model.Based on this,a method of information extraction based on neural network is proposed.The corresponding models are established from the aspects of face information extraction and action information extraction,and the experimental and analysis results are given.Experimental results show that this method can extract information effectively and has high recognition accuracy.