1 Introduction

Title : Summary Level Training of Sentence Rewriting for Abstractive Summarization
Link : http://arxiv.org/abs/1909.08752
Author : Sanghwan Bae, Taeuk Kim, Jihoon Kim and Sang-goo Lee
Conference : Proceeding of the 2nd Workshop on New Frontiers in Summarization (held by EMNLP)

1.1 Achievement

  1. Present a novel training signal that directly maximizes summary-level ROUGE scores through reinforcement learning
  2. Obtains new state-of-the-art performance on both CNN/Daily Mail and New York Times datasets[1]

[1] Actually, the result of (Liu, 2019) is now state-of-the-art.

2 Method

2.1 Extractor

Requirement : Input with document sequence D and output with the chosen extractive sentence subset

Solution : (Sequence D -> BERTSUM (Encoder) -> LSTM Pointer Network (Decoder))

  1. Input the sentence vector into BERTSUM[1] in which they use BERT BASE, a small version of BERT LARGE
  2. Feed the sentence representation vector H into the LSTM Pointer Network[2] with 256 hidden size

Advantage :

  1. Enable the decoder to consider the previous selected sentence
  2. Avoid information overlapping between the selected sentences

[1] Text Summarization with Pretrained Encoder, 2019
[2] Pointer networks, 2015

2.2 Abstractor[1]

Requirement : Input with selected extractive sentence subset and output with the corresponding paraphased sentence

Solution : (extractive sentence subset -> seq2seq model)

  1. Use word2vec[2] model of 128 dimensions
  2. Feed extractive sentence subset modified by word2vec into seq2seq[3] model which use copying mechanism[4] for handling out-of-vocabulary words

[1] Fast Abstractive Summarization with Reinforce-Selected Sentence Rewriting, 2018
[2] Distributed representations of words and phrases and their compositionality, 2013
[3] Neural machine translation by jointly learning to align and translate, 2015; Effective approaches to attention-based neural machine translation, 2015
[4] Get to the point: Summarization with pointer-generator networks, 2017

3 Training

3.1 Extractor Pre-training[1]

Requirement : Selected sentences should be the ones that maximize the Rouge score with respect to gold summaries

Problem :

  1. Most summarization corpora only contain human written abstractive summaries as ground truth (No extractive summary for training)
  2. It is computationally expensive to find a globally optimal subset of sentences that maximizes the Rouge score

Solution :

  1. Add the document sentence to the extractive oracles until the remaining candidate can’t improve the Rouge score with respect to entire gold summary
  2. Train the extractive model with cross-entropy

[1] Summarunner: A recurrent neural network based sequence model for extractive summarization of documents, 2017

3.2 Abstractor Training

Requirement : Find the proper paraphrased function between original sentence and the gold summary

Solution :

  1. For each ground truth summary sentence, find the most similar sentence in the original document with Rouge score
  2. Train the model as a usual seq2seq model with cross-entropy

3.3 Reinforcement Learning Optimization

Requirement : Optimize the Rouge metric

Solution : Assume the extractor as an agent in reinforcement learning paradigm

Not quite understand yet.

4 Result

Their Rouge score is worse than the score of (Liu, 2019) .

5 Comparison and Conclusion

In this paper :

  1. Extractor : BERTSUM (Encoder) + LSTM (Decoder)
  2. Abstractor : seq2seq

In (Liu, 2019) :

  1. Extractor : BERTSUM (Encoder) + Transformer + sigmoid classifier (Decoder)
  2. Abstractor : 6 layered Transformer

Conclusion : Seems that Transformer in (Liu, 2019) performs better than seq2seq in this paper

喵喵喵？