Closed-Book Training to Improve Summarization Encoder Memory
arXiv:1809.04585
Abstract
A good neural sequence-to-sequence summarization model should have a strong encoder that can distill and memorize the important information from long input texts so that the decoder can generate salient summaries based on the encoder's memory. In this paper, we aim to improve the memorization capabilities of the encoder of a pointer-generator model by adding an additional 'closed-book' decoder without attention and pointer mechanisms. Such a decoder forces the encoder to be more selective in the information encoded in its memory state because the decoder can't rely on the extra information provided by the attention and possibly copy modules, and hence improves the entire model. On the CNN/Daily Mail dataset, our 2-decoder model outperforms the baseline significantly in terms of ROUGE and METEOR metrics, for both cross-entropy and reinforced setups (and on human evaluation). Moreover, our model also achieves higher scores in a test-only DUC-2002 generalizability setup. We further present a memory ability test, two saliency metrics, as well as several sanity-check ablations (based on fixed-encoder, gradient-flow cut, and model capacity) to prove that the encoder of our 2-decoder model does in fact learn stronger memory representations than the baseline encoder.
arXiv commentEMNLP 2018 (16 pages)
References in corpus (15)
- Sequence to Sequence Learning with Neural Networks
- Teaching Machines to Read and Comprehend
- Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks
- A Deep Reinforced Model for Abstractive Summarization
- SQuAD: 100,000+ Questions for Machine Comprehension of Text
- Improved Image Captioning via Policy Gradient optimization of SPIDEr
- Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyond
- Incorporating Copying Mechanism in Sequence-to-Sequence Learning
- An Actor-Critic Algorithm for Sequence Prediction
- A Sentence Compression Based Framework to Query-Focused Multi-Document Summarization
- Reward Augmented Maximum Likelihood for Neural Structured Prediction
- Efficient Summarization with Read-Again and Copy Mechanism
- Pointing the Unknown Words
- Language as a Latent Variable: Discrete Generative Models for Sentence Compression
- Memory-enhanced Decoder for Neural Machine Translation