Generating Multiple-Length Summaries via Reinforcement Learning for Unsupervised Sentence Summarization
This addresses the problem of flexible summarization without human-written data for NLP applications, though it is incremental as it builds on existing reinforcement learning and abstractive methods.
The paper tackles unsupervised abstractive sentence summarization by developing a reinforcement learning model that generates multiple summaries of varying lengths, achieving substantial performance improvements over both abstractive and extractive models.
Sentence summarization shortens given texts while maintaining core contents of the texts. Unsupervised approaches have been studied to summarize texts without human-written summaries. However, recent unsupervised models are extractive, which remove words from texts and thus they are less flexible than abstractive summarization. In this work, we devise an abstractive model based on reinforcement learning without ground-truth summaries. We formulate the unsupervised summarization based on the Markov decision process with rewards representing the summary quality. To further enhance the summary quality, we develop a multi-summary learning mechanism that generates multiple summaries with varying lengths for a given text, while making the summaries mutually enhance each other. Experimental results show that the proposed model substantially outperforms both abstractive and extractive models, yet frequently generating new words not contained in input texts.