Research on Intelligent Portrait of Chinese Elderly Based on Big Data and Deep Learning

The growth rate of the elderly population in China has accelerated. According to UN standards, China has fully entered the aging society. In the future, the trend of population aging is irreversible. At the same time, Internet and mobile technologies are developing rapidly. We can use it to actively and effectively respond to the aging of China's population. Based on the data source under the health media of the elderly, this paper constructs the portraits of the elderly, focusing on the different aspects of the elderly, including clothing, food, housing and other aspects, and classifying them. This paper, extracts the local features of the text by constructing a word embedding sequence and imports it into the bidirectional Long Short-Term Memory ( LSTM) model. Through the attention mechanism and the LSTM model combined with the text importance calculation, the model is obtained by classifying the overall characteristics of the text. At the same time, a deep convolutional neural network model based on word embedding and multiple convolution kernels is proposed to deal with text multi-classification problem. In order to verify the accuracy of the model, 8000 old people's state and behavioral habit data were used in the experiment. The experimental results show that the model constructed in this paper is more accurate than the traditional machine learning model and the standard deep learning model.