Text abstract generation system and method based on adversarial learning and hierarchical neural network

Abstract

The invention requests to protect a text abstract generation system and method based on adversarial learning and a hierarchical neural network, and belongs to the field of text abstracts of natural language processing. The system comprises a discriminator module, a preprocessing module, a word embedding module, a sentence embedding module, a generation module and an adversarial learning module. According to the invention, on the basis of an encoder decoder model (Seq2Seq), a new hierarchical division model is provided. An encoder part of the Seq2Seq is divided into a word embedding layer and asentence embedding layer, and an enhanced memory mechanism is introduced into each layer, so that the model can better understand text meanings, adversarial learning is introduced during decoding, arecognizer is arranged to recognize standard representation and fuzzy representation, the distance between the standard representation and the fuzzy representation is shortened, and meanwhile, learning is supervised to prevent the standard representation and the fuzzy representation from approaching, confrontation is formed, and when confrontation is balanced, an optimal generation result is found, so that the text abstract generation accuracy is improved.

Description

technical field [0001] The invention belongs to the field of text summarization of natural language processing, in particular to a text summarization method and system based on adversarial learning and layered neural network. Background technique [0002] With the explosive growth of Internet text information in recent years, people are exposed to massive amounts of text information every day, such as news, blogs, chats, reports, papers, Weibo, etc. Extracting important content from a large amount of text information has become an urgent need, and automatic text summarization provides an efficient solution. [0003] The traditional attention mechanism-based encoder-decoder model first encodes the words of the text, then adds the attention mechanism to learn the key information of the article, and then decodes the word encoding to generate a text summary. The granularity of the attention mechanism of this type of method is relatively coarse, and the learning of long texts ca...

AI Technical Summary

Problems solved by technology

For the traditional seq2seq model based on the attention mechanism, the input sequence is too long to memorize the context well, resulting in the lack of text understanding and the coarser granularity of the attention mechanism, which leads to inaccurate capture of key information and low summary accuracy. For the problem, an idea of ​​layered encoding of neural network is proposed, which is divided into word embedding level and sentence embedding level, and the enhanced memory mechanism is introduced at each level. The benefit is to reduce the error in backpropagation derivation and refine the The attention granularity of the attention mechanism of the traditional seq2seq model enables it to more accurately capture the key information in the article. At the same time, it introduces adversarial learning when decoding, and sets up a recognizer to identify the standard representation and the fuzzy representation, reducing the gap between the two. At the same time, supervised learning prevents them from approaching, forming a confrontation, and finding the optimal generation result when the confrontation is balanced, improving the accuracy of text summary generation, thereby improving the accuracy of the final generated summary

Images