在漫长的马尔科夫链设想的笼罩下，下一个词仅仅取决于上一个的设想，这跟实际情况是不符合的。而RNN的出现让这种情形有所缓解。 相关论文有：

• Incorporating Copying Mechanism in Sequence-to-Sequence Learning

• Controlling Output Length in Neural Encoder-Decoders

• Online Segment to Segment Neural Transduction

• Topic Sensitive Neural Headline Generation

• Selective Encoding for Abstractive Sentence Summarization

• Faithful to the Original: Fact Aware Neural Abstractive Summarization

• Neural Headline Generation with Sentence-wise Optimization

• Deep Recurrent Generative Decoder for Abstractive Text Summarization

• A Reinforced Topic-Aware Convolutional Sequence-to-Sequence Model for Abstractive Text Summarization

• Generating News Headlines with Recurrent Neural Networks