融合上下文信息和关键信息的文本摘要  

Fusing Context Information and Key Information for Text Summarization

作　　者：李志欣[1] 彭智 唐素勤[1] 马慧芳 LI Zhixin;PENG Zhi;TANG Suqin;MA Huifang(Guangxi Key Lab of Multi-source Information Mining and Security,Guangxi Normal University,Guilin,Guangxi 541004,China;School of Computer Science and Engineering,Northwest Normal University,Lanzhou,Gansu 730070,China)

机构地区：[1]广西师范大学广西多源信息挖掘与安全重点实验室,广西桂林541004 [2]西北师范大学计算机科学与工程学院,甘肃兰州730070

出　　处：《中文信息学报》2022年第1期83-91,共9页Journal of Chinese Information Processing

基　　金：国家自然科学基金(61966004,61663004,61967002,61866004,61762078);广西自然科学基金(2019GXNS-FDA245018,2018GXNSFDA281009);广西八桂学者创新科研团队项目。

摘　　要：的一个迫切需要解决的问题是如何准确地概括文本的核心内容。目前文本摘要的主要方法是使用编码器-解码器架构,在解码过程中利用软注意力获取所需的上下文语义信息。但是,由于编码器有时候会编码过多的信息,所以生成的摘要不一定会概括源文本的核心内容。为此,该文提出一种基于双注意指针网络的文本摘要模型。首先,该模型使用了双注意指针融合网络,其中自注意机制从编码器中收集关键信息,软注意和指针网络通过上下文信息生成更连贯的核心内容。两者融合能够生成具有总结性和连贯性的摘要。其次,采用改进后的覆盖率机制来处理重复问题,提高生成摘要的准确性。同时,结合计划采样和强化学习产生新的训练方法来优化模型。在CNN/Daily Mail数据集和LCSTS数据集上的实验表明,该模型达到了当前主流模型的效果。实验结果分析表明,该模型在总结性方面具有良好的表现,同时减少了重复的出现。In text summarization, the mainstream method is to use encoder-decoder architecture to obtain the required context semantic information by using soft attention in the decoding process. Since the encoder sometimes encodes too much information, the generated summary does not always summarize the core content of the source text. To address this issue, this paper proposes a text summarization model based on a dual-attention pointer network. Firstly, in the dual-attention pointer network, the self-attention mechanism collects key information from the encoder, while the soft attention and the pointer network generate more coherent core content through context information. The fusion of both will generate accurate and coherent summaries. Secondly, the improved coverage mechanism is applied to address the repetition problem and improve the quality of the generated summaries. Simultaneously, scheduled sampling and reinforcement learning are combined to generate new training methods to optimize the model. Experiments on the CNN/Daily Mail dataset and the LCSTS dataset show that the proposed model performs as well as many state-of-the-art models.

关 键 词：文本摘要 神经网络 注意力机制 指针网络 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]