Automatic text summarization(ATS)has achieved impressive performance thanks to recent advances in deep learning(DL)and the availability of large-scale corpora.The key points in ATS are to estimate the salience of info...Automatic text summarization(ATS)has achieved impressive performance thanks to recent advances in deep learning(DL)and the availability of large-scale corpora.The key points in ATS are to estimate the salience of information and to generate coherent results.Recently,a variety of DL-based approaches have been developed for better considering these two aspects.However,there is still a lack of comprehensive literature review for DL-based ATS approaches.The aim of this paper is to comprehensively review significant DL-based approaches that have been proposed in the literature with respect to the notion of generic ATS tasks and provide a walk-through of their evolution.We first give an overview of ATS and DL.The comparisons of the datasets are also given,which are commonly used for model training,validation,and evaluation.Then we summarize single-document summarization approaches.After that,an overview of multi-document summarization approaches is given.We further analyze the performance of the popular ATS models on common datasets.Various popular approaches can be employed for different ATS tasks.Finally,we propose potential research directions in this fast-growing field.We hope this exploration can provide new insights into future research of DL-based ATS.展开更多
法律文本的自动生成能缓解我国法律服务行业中的人力资源不足的问题,对抗生成网络模型的出现为法律文本的自动生成提供了新思路.本文提出一种基于对抗生成网络的文本自动生成模型——ED-GAN(Generative Adversarial Networks based on E...法律文本的自动生成能缓解我国法律服务行业中的人力资源不足的问题,对抗生成网络模型的出现为法律文本的自动生成提供了新思路.本文提出一种基于对抗生成网络的文本自动生成模型——ED-GAN(Generative Adversarial Networks based on Encoder-Decoder).在该模型的生成器中,首先将案情要素的关键词序列输入至编码器Encoder阶段的LSTM中编码成一隐含层向量,再将这个隐含层向量输入到解码器Decoder的LSTM中,并结合其各时间步的输出生成下一时间步的隐含层向量,进而得到各时间步的输出,生成文本序列.模型最后采用CNN网络来鉴别生成文本和真实文本之间的差距.实验验证表明,采用所提模型能够生成较理想的法律文本.展开更多
The satellite-based automatic identification system (AIS) receiver has to encounter the frequency offset caused by the Doppler effect and the oscillator instability. This paper proposes a non-coherent sequence detecti...The satellite-based automatic identification system (AIS) receiver has to encounter the frequency offset caused by the Doppler effect and the oscillator instability. This paper proposes a non-coherent sequence detection scheme for the satellite-based AIS signal transmitted over the white Gaussian noise channel. Based on the maximum likelihood estimation and a Viterbi decoder, the proposed scheme is capable of tolerating a frequency offset up to 5% of the symbol rate. The complexity of the proposed scheme is reduced by the state-complexity reduction, which is based on per-survivor processing. Simulation results prove that the proposed non-coherent sequence detection scheme has high robustness to frequency offset compared to the relative scheme when messages collision exists.展开更多
基金supported by the National Key Research and Development Program of China under Grant No.2016YFB1000902the National Natural Science Foundation of China under Grant Nos.61232015,61472412,and 61621003.
文摘Automatic text summarization(ATS)has achieved impressive performance thanks to recent advances in deep learning(DL)and the availability of large-scale corpora.The key points in ATS are to estimate the salience of information and to generate coherent results.Recently,a variety of DL-based approaches have been developed for better considering these two aspects.However,there is still a lack of comprehensive literature review for DL-based ATS approaches.The aim of this paper is to comprehensively review significant DL-based approaches that have been proposed in the literature with respect to the notion of generic ATS tasks and provide a walk-through of their evolution.We first give an overview of ATS and DL.The comparisons of the datasets are also given,which are commonly used for model training,validation,and evaluation.Then we summarize single-document summarization approaches.After that,an overview of multi-document summarization approaches is given.We further analyze the performance of the popular ATS models on common datasets.Various popular approaches can be employed for different ATS tasks.Finally,we propose potential research directions in this fast-growing field.We hope this exploration can provide new insights into future research of DL-based ATS.
文摘法律文本的自动生成能缓解我国法律服务行业中的人力资源不足的问题,对抗生成网络模型的出现为法律文本的自动生成提供了新思路.本文提出一种基于对抗生成网络的文本自动生成模型——ED-GAN(Generative Adversarial Networks based on Encoder-Decoder).在该模型的生成器中,首先将案情要素的关键词序列输入至编码器Encoder阶段的LSTM中编码成一隐含层向量,再将这个隐含层向量输入到解码器Decoder的LSTM中,并结合其各时间步的输出生成下一时间步的隐含层向量,进而得到各时间步的输出,生成文本序列.模型最后采用CNN网络来鉴别生成文本和真实文本之间的差距.实验验证表明,采用所提模型能够生成较理想的法律文本.
文摘The satellite-based automatic identification system (AIS) receiver has to encounter the frequency offset caused by the Doppler effect and the oscillator instability. This paper proposes a non-coherent sequence detection scheme for the satellite-based AIS signal transmitted over the white Gaussian noise channel. Based on the maximum likelihood estimation and a Viterbi decoder, the proposed scheme is capable of tolerating a frequency offset up to 5% of the symbol rate. The complexity of the proposed scheme is reduced by the state-complexity reduction, which is based on per-survivor processing. Simulation results prove that the proposed non-coherent sequence detection scheme has high robustness to frequency offset compared to the relative scheme when messages collision exists.
