GEMINI: Controlling the Sentence-level Writing Style for Abstractive Text Summarization
This work addresses the problem of flexible sentence-level writing style control in abstractive text summarization for NLP researchers, representing an incremental improvement over existing methods.
The paper tackles the challenge of imitating human summarization techniques by proposing GEMINI, an adaptive model that integrates rewriting and generating to produce summaries, which outperforms baselines on three datasets and achieves the best results on WikiHow.
Human experts write summaries using different techniques, including extracting a sentence from the document and rewriting it, or fusing various information from the document to abstract it. These techniques are flexible and thus difficult to be imitated by any single method. To address this issue, we propose an adaptive model, GEMINI, that integrates a rewriter and a generator to mimic the sentence rewriting and abstracting techniques, respectively. GEMINI adaptively chooses to rewrite a specific document sentence or generate a summary sentence from scratch. Experiments demonstrate that our adaptive approach outperforms the pure abstractive and rewriting baselines on three benchmark datasets, achieving the best results on WikiHow. Interestingly, empirical results show that the human summary styles of summary sentences are consistently predictable given their context. We release our code and model at \url{https://github.com/baoguangsheng/gemini}.