一种基于实体的邮件任务提取策略TEXEM：一种基于实体的.pdfVIP

TEXEM ：一种基于实体的邮件任务提取策略 TEXEM ：一种基于实体的邮件任务提取策略 1 2 1 1 张相於 陈继东 李玉坤 孟小峰 1 （中国人民大学信息学院，北京 100872 ） 2 （EMC 中国实验室，北京 100084 ） （zhangxy@ ） 摘要 在信息化飞速发展的今天，电子邮件的使用正在变得越来越频繁，而且其应用场合也在不断扩展，目前世界上很大一部 分、并且越来越多的商业和个人往来都是通过电子邮件完成的。电子邮件的作用正在被人们不断拓展，特别是它作为任务管理 和协作管理的重要工具和手段，被人们广泛使用。但是由于电子邮件数量越来越大，其设计初衷只是单纯的通讯手段，具有轻 量性和随意性等特点，因此当前的邮件管理工具很难有效的组织和管理人们的大量邮件信息，更重要的是它们不能很好的完成 邮件任务管理等扩展任务。本文提出了一种基于实体的邮件任务提取框架——TEXEM ，充分考虑电子邮件中的结构信息，将邮 件通过实体聚类的方法转化为任务的集合，并对用户任务重要性进行评估，使用户不再单纯依靠毫无结构的纯文本内容来处理 邮件，以起到辅助用户邮件处理过程、提高邮件管理效率的作用。 关键词 任务提取；实体识别；邮件处理；聚类 中图法分类号 TP393 TEXEM ： An Entity-based Task Extraction Approach for Emails 1 2 1 1 Zhang Xiangyu , Chen Jidong , Li Yukun , Meng Xiaofeng 1 (School of Information, Renmin University of China, Beijing 100872) 2 (EMC Research China, Beijing 100084) Abstract With the development of information technology, usage of email is getting more and more popular and intense. Usage of email has been extended to many scenarios. Nowadays, a large portion of commercial and personal communication is delivered through emails, and the number never stops growing. As an important information repository, email is expected to be much more than a simple communication channel. In particular, people are using email to perform tasks management and collaboration management. However, since email has large volume and is designed as a communications application, it’s born with characteristics of lightweight and casualness Using current email management tools, people have difficulties in organizing and managing their email data. Most importantly they have problems in using email to perform tasks management. In this paper, we propose a framework of entity-based task extraction from emails