In this work, we define a new style transfer task: perspective shift, which reframes a dialogue from informal first person to a formal third person rephrasing of the text. This task requires challenging coreference resolution, emotion attribution, and interpretation of informal text. We explore several baseline approaches and discuss further directions on this task when applied to short dialogues. As a sample application, we demonstrate that applying perspective shifting to a dialogue summarization dataset (SAMSum) substantially improves the zero-shot performance of extractive news summarization models on this data. Additionally, supervised extractive models perform better when trained on perspective shifted data than on the original dialogues. We release our code publicly.

翻译：在这项工作中,我们定义了一种新的风格传输任务:视角转换,将对话从非正式的第一人转变为正式的第三人,对文本进行修改。这项任务要求具有挑战性的共同参照分辨率、情感归属和对非正式文本的解释。我们探索了几种基线方法,并在应用到短期对话时讨论了关于这项任务的进一步方向。作为样本应用,我们证明将视角应用到对话汇总数据集(SAMSum)可大大改善这一数据上的采掘新闻汇总模型的零光性能。此外,监督的采掘模型在接受关于视角转换数据的培训时比在原始对话方面表现更好。我们公开发布我们的代码。