Reinforced Generative Adversarial Network for Abstractive Text Summarization
This addresses limitations in abstractive text summarization for researchers needing accurate, non-repetitive summaries, though it appears incremental as it builds on existing sequence-to-sequence and adversarial approaches.
The authors tackled the problem of abstractive text summarization by proposing a reinforced generative adversarial network that combines reinforcement learning with adversarial networks to enhance sequence-to-sequence models. They achieved competitive results on a COVID-19 paper title summarization task, matching current models on ROUGE scores while improving readability.
Sequence-to-sequence models provide a viable new approach to generative summarization, allowing models that are no longer limited to simply selecting and recombining sentences from the original text. However, these models have three drawbacks: their grasp of the details of the original text is often inaccurate, and the text generated by such models often has repetitions, while it is difficult to handle words that are beyond the word list. In this paper, we propose a new architecture that combines reinforcement learning and adversarial generative networks to enhance the sequence-to-sequence attention model. First, we use a hybrid pointer-generator network that copies words directly from the source text, contributing to accurate reproduction of information without sacrificing the ability of generators to generate new words. Second, we use both intra-temporal and intra-decoder attention to penalize summarized content and thus discourage repetition. We apply our model to our own proposed COVID-19 paper title summarization task and achieve close approximations to the current model on ROUEG, while bringing better readability.