论文解读 - SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive Summarization of Documents
阅读数:
999+次
字数统计 : 1181
阅读时间 ≈ 3.0 分钟
发布时间 : 2019-03-25
Nallapati R, Zhai F, Zhou B. Summarunner: A recurrent neural network based sequence model for extractive summarization of documents[C]. Thirty-First AAAI Conference on Artificial Intelligence. 2017.
the relative position refers to a quantized representation that divides each document into a fixed number of segments and computes the segment ID of a given sentence.
Intuitively, since the summary representation acts as the only information channel between the SummaRuNNer model and the decoder, maximizing the probability of abstractive summary words as computed by the decoder will require the model to learn a good summary representation which in turn depends on accurate estimates of extractive probabilities p(yj).