Research on the Application of Deep Learning in Text Generation

Abstract

In recent years, deep learning technology has made great progress, and text generation based on deep learning has received extensive attention. In the form of complex and diverse information, text information is a mainstream form of data, and its number is growing very fast. How to quickly and accurately locate and use effective information from massive text data has become an urgent research problem in the field of text information extraction. The deepening of deep learning technology has broken through the dependence of traditional natural language generation technology on templates. It can automatically learn the input to output mapping from the data to form an end-to-end solution and reduce the degree of human participation. It enables the generation system to generalize in a wider field, and can generate more free text under the given conditions according to the needs. This paper explores the text generation method based on deep neural network, conducts research work on the text summary task, designs a generative summary generation method based on improved cluster search, and conducts experiments. Experimental results show that this method is effective.