【24h】

Statistical Framework for Email Discourse Analysis

机译:电子邮件语篇分析的统计框架

获取原文
获取原文并翻译 | 示例

摘要

The discourse structure of emails is an important but largely unaddressed problem. Emails or newsgroup messages are usually organized as threads - a strict tree structure with the original message on the top and the responses at the bottom of the tree. Such a tree models the message-response relationship as a parent-child relationship. In addition to the parent-child relationship between an email and response, we claim that a rhetorical relation also exists between them. These rhetorical relations reflect the nature of exchanges and can be used in tasks like thread summarization. In this paper, we present a framework for Email Discourse Analysis and also an algorithm that learns to characterize the rhetorical relationships in an email thread. We present the results of our algorithm on a collection of messages obtained from the newsgroups.
机译:电子邮件的话语结构是一个重要但基本上未解决的问题。电子邮件或新闻组消息通常组织为线程-一种严格的树结构,原始消息位于顶部,响应位于树的底部。这样的树将消息-响应关系建模为父子关系。除了电子邮件和回复之间的父子关系外,我们还声称它们之间也存在修辞关系。这些修辞关系反映了交换的性质,可用于线程摘要之类的任务。在本文中,我们提出了电子邮件话语分析的框架,以及一种学习表征电子邮件线程中修辞关系的算法。我们在从新闻组获得的消息集合上展示算法的结果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号