Efficient Summarization with Read-Again and Copy Mechanism
This addresses efficiency and accuracy issues in text summarization for NLP applications, representing an incremental improvement.
The paper tackled suboptimal word representations and slow decoding in encoder-decoder models by introducing a read-again mechanism and a copy mechanism, achieving state-of-the-art results on the Gigaword dataset and DUC competition.
Encoder-decoder models have been widely used to solve sequence to sequence prediction tasks. However current approaches suffer from two shortcomings. First, the encoders compute a representation of each word taking into account only the history of the words it has read so far, yielding suboptimal representations. Second, current decoders utilize large vocabularies in order to minimize the problem of unknown words, resulting in slow decoding times. In this paper we address both shortcomings. Towards this goal, we first introduce a simple mechanism that first reads the input sequence before committing to a representation of each word. Furthermore, we propose a simple copy mechanism that is able to exploit very small vocabularies and handle out-of-vocabulary words. We demonstrate the effectiveness of our approach on the Gigaword dataset and DUC competition outperforming the state-of-the-art.