论文标题
抽象文本摘要中的多事实校正
Multi-Fact Correction in Abstractive Text Summarization
论文作者
论文摘要
预先训练的神经抽象摘要系统至少在Rouge方面已经主导了有关新闻摘要绩效的提取策略。但是，系统生成的抽象摘要通常面临事实不一致的陷阱：就源文本产生错误的事实。为了应对这一挑战，我们提出了Span-Fact，这是两个事实校正模型组成的套件，这些模型利用了从问题回答模型中学到的知识，以通过SPAN选择在系统生成的摘要中进行更正。我们的模型采用单一或多掩膜策略来迭代或自动回归替换实体，以确保语义一致性W.R.T.源文本同时保留了通过抽象摘要模型生成的摘要的句法结构。实验表明，我们的模型显着提高了系统生成的摘要的事实一致性，而无需牺牲自动指标和人类评估的摘要质量。
Pre-trained neural abstractive summarization systems have dominated extractive strategies on news summarization performance, at least in terms of ROUGE. However, system-generated abstractive summaries often face the pitfall of factual inconsistency: generating incorrect facts with respect to the source text. To address this challenge, we propose Span-Fact, a suite of two factual correction models that leverages knowledge learned from question answering models to make corrections in system-generated summaries via span selection. Our models employ single or multi-masking strategies to either iteratively or auto-regressively replace entities in order to ensure semantic consistency w.r.t. the source text, while retaining the syntactic structure of summaries generated by abstractive summarization models. Experiments show that our models significantly boost the factual consistency of system-generated summaries without sacrificing summary quality in terms of both automatic metrics and human evaluation.
