Research On Machine Intelligence Writing Technology Based On NLP

With the progress and development of The Times,as well as the technological revolution of the integration of big data,cloud computing and Internet of things,natural language processing technology has made rapid development in a certain field,especially playing a very critical role in intelligent writing.Based on double-byte languages especially Chinese characters in Chinese in the asia-pacific region as the research object,based on the theory of natural language processing,try to establish intelligent writing calculation model,and let the machine by calculation of natural language to convey ideas and meaning,then make a reasonable response,to explore the quantitative relationship between the text and explore further reveal the mechanism of natural language processing,machine understanding the laws of natural language.Based on the systematic study of NLP technology and the exploration of the law of automatic text generation model,this paper proposes to construct a language model based on RNN two-layer network algorithm,which provides theoretical basis for intelligent writing and technical support for the automatic text generation.The specific research content includes the following three aspects:1.By analyzing and studying the current technologies of natural language processing and machine learning,this paper explores the rules of automatic text generation model.Based on the research of the application of natural language processing technology in Chinese text and the method of machine learning--deep learning,this paper focuses on the research of text generation and implementation.2.Build a language model based on RNNIn understanding the text generation technique and based on NLP technology,based on RNN double network language model is put forward,the model with the method of increasing network layer,a layer of the network to replace the original solve the problems of the neural network weight too much,at the same time with the length of memory model structure to replace the original RNN structure to solve the problem of sequence text generated rely on for a long time.This model is based on the deep learning method.By improving the traditional cyclic neural network algorithm,the Chinese text information is processed and the language model is constructed with the TensorFlow machine learning tool.3.An intelligent writing system based on NLP is implemented by RNN double layer network algorithmIn order to verify and evaluate the RNN-based two-layer network language model constructed in this paper,an intelligent writing poetry design system was designed and implemented.The system uses more than 40,000 Tang poems as a data set.After cleaning the data set,by calling the trained language model,finally,inputting a keyword to the system allows the machine to intelligently create a complete poem.In order to make the generated poetry more artistic,the PIL library is used to integrate the visual display effect into the system,and the generated poems are embedded in the image display.