arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.12727cs.CL
收藏

基于潜在叙事结构的电影剧本摘要

Screenplay Summarization Using Latent Narrative Structure

University of Edinburgh(爱丁堡大学);University of Melbourne(墨尔本大学)

Pinelopi Papalampidi, Frank Keller, Lea Frermann, Mirella Lapata

摘要:

大多数通用提取摘要模型都是在新闻文章上训练的,这些文章很短,并且预先呈现了所有重要信息。因此,这样的模型在位置上是有偏见的,并且通常从文档的开头开始执行句子的智能选择。当总结长篇叙事时,它具有复杂的结构和零碎的信息,简单的位置分析是不够的。在本文中,我们建议将叙事的底层结构明确纳入一般的无监督和监督提取摘要模型。我们根据关键叙事事件(转折点)将叙事结构形式化,并将其视为潜在的,以总结剧本(即,提取场景的最佳序列)。CSI语料库的电视剧本,我们增加了场景级的摘要标签的实验结果表明,潜在的转折点与CSI情节的重要方面,并提高概括性能一般提取算法,导致更完整和多样化的摘要。

英文摘要:

Most general-purpose extractive summarization models are trained on news articles, which are short and present all important information upfront. As a result, such models are biased on position and often perform a smart selection of sentences from the beginning of the document. When summarizing long narratives, which have complex structure and present information piecemeal, simple position heuristics are not sufficient. In this paper, we propose to explicitly incorporate the underlying structure of narratives into general unsupervised and supervised extractive summarization models. We formalize narrative structure in terms of key narrative events (turning points) and treat it as latent in order to summarize screenplays (i.e., extract an optimal sequence of scenes). Experimental results on the CSI corpus of TV screenplays, which we augment with scene-level summarization labels, show that latent turning points correlate with important aspects of a CSI episode and improve summarization performance over general extractive algorithms leading to more complete and diverse summaries.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑