文本摘要简述.pdf

文本摘要简述 Xiachong Feng 目录 • 简介 • 分类 • 方法 • 评价指标 • 数据集 • 总结 简介 • 文本摘要旨在将文本或文本集合转换为包含关 键信息的简短文本。 分类 • 输入类型 • 单文档摘要(Single-Document) • 多文档摘要(Multi-Document) • 输出类型 • 抽取式摘要(Extractive) • 生成式摘要(Abstractive) • 学习方法 • 有监督摘要(Supervised) • 无监督摘要(Unsupervised) • 语言种类 • 单语言摘要(Monolingual) • 跨语言摘要(Cross lingual) • …… 1. 减小句子冗余度 2. 确定句子顺序 单文档 && 多文档 3. 确定压缩比率 Topic “Malaysia Airlines Disappearance” Summary Flight MH370, carrying 239 people vanished over the South China Sea in less than an hour after taking off from Kuala Lumpur, with two passengers boarded the Boeing 777 using stolen passports. Possible reasons could be an abrupt breakup of the plane or an act of terrorism…… 抽取式 && 生成式 抽取式摘要 生成式摘要 Nallapati, R., Zhou, B., santos, dos, C. N., Gülçehre, Ç., & Xiang, B. (2016, February 19). Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyond. arX. 单语言 && 跨语言 • Cross-Language Summarization (*-2018) • Cross-Lingual Summarization (2019) Source Language summarization Target Language Document Summary (e.g. English) (e.g. Chinese