Research And Implementation Of Intelligent Reading Comprehension Problem Solving System For Elementary School Student

With the rise of the Internet,more and more data has been accumulated.At the same time,Deep neural network models have gained unprecedented progress in many places. Reading comprehension is an important mission of Natural Language Processing that more and more people began to try to solve it by the deep neural network model.A simple description of the reading comprehension for the machine is to read a text, and then answer the question according to the text. With the widespread use of artificial intelligence in people’s lives, people look forward to another day that the machine can also be able to read the text, and understand the meaning of it. And eventually that can help people to read and find useful information in the text, of course that information must be accurate information. In this paper, we use the traditional methods of natural processing and deep neural network to deal with the different tasks of reading comprehension. To apply the deep neural network must rely on big data. So if we want to train the neural network, we have to obtain large training data. In the small data set,we can not fully use the neural network’s huge learning ability. This is the reason why the neural network was not very practical when it came out. we combine the manual labeling and the weak rules to construct the training data. In this way, we can get massive training data. In the selection of neural network, we use a Attention based neural network model. It simulates people’s reading habits,when reading it can automatically find the important part of the content. The key content is generally used to answer the question of reading comprehension. In the application, we also found that the learning ability of Attention based neural network model in different reading comprehension data set is higher than that of conventional deep neural networks, this point also have been verified in the field of machine translation, image recognition and so on.Finally, we construct a set of reading comprehension problem solving system by combining the conventional Natural Language Processing technology and the deep learning neural network model. In this system, we can show the latest research content and results in real time, and verify the validity of the neural network model.And it can also visually display the whole system of problem-solving process.