利用分割技术提高对话文本自动摘要的一致性

针对传统的基于抽取的自动摘要方法不能很好地处理对话文本一致性的问题,提出了对话文本自动摘要中的体裁、问题-回答对识别等关键技术.设计了一种类似于TextTiling的文本分割方法对候选的回答段落进行分割,以便提高问题-回答对的关联精度.实验结果表明,利用问题-回答对的关联信息能够提高对话文本自动摘要的整体一致性.

Using Text Segmentation to Improve the Coherence of Chinese Dialogue Text Summary

Automatic text summarization for dialogue style is a relatively new research area.To reduce the un-coherence of the summary when the conventional extract-based automatic text summarization approaches for the no-dialogue documents are applied to the dialogue documents,some key techniques of automatic Chinese text summarization for dialogue style were proposed,such as identifying document's style,identifying the questions and their corresponding answers and linking them to form question-answer pairs.An automatic text segmentation approach analogous to TextTiling was exploited to improve the precision of correlating question paragraphs and answer paragraphs.The experimental results show that the approach improves significantly the global coherence of the summary.