计算机工程与应用 ›› 2015, Vol. 51 ›› Issue (19): 141-145.

• 数据库、数据挖掘、机器学习 • 上一篇    下一篇

交互式问答中基于话语结构的指代消解研究

张  超,孔  芳,周国栋   

  1. 苏州大学 计算机科学与技术学院 自然语言处理实验室,江苏 苏州 215006
  • 出版日期:2015-09-30 发布日期:2015-10-13

Coreference resolution in IQA based on discourse structure

ZHANG Chao, KONG Fang, ZHOU Guodong   

  1. NLP Lab, School of Computer Science and Technology, Soochow University, Suzhou, Jiangsu 215006, China
  • Online:2015-09-30 Published:2015-10-13

摘要: 交互式问答系统能够与用户进行对话式交互进而处理用户提出的一系列问题,但指代现象阻碍了系统准确地理解用户的问题。深入研究了交互式问答中的指代消解任务,在交互式问答文本语料中标注指代链信息和话语结构信息,并在基准平台的基础上提出了基于话语结构的特征集。实验研究了新闻文本上提出的基准特征集在交互式问答文本中的性能,并在此基础上引入话语结构特征。实验结果表明,与基准系统相比,改进系统的[F]值提高了2.6%,指代消解平台的性能有较大的提升。

关键词: 交互式问答, 话语结构, 指代消解

Abstract: Interactive Question Answering(IQA), a hot research topic in the area of QA, allows system to interact with users to process a series of questions just with a dialogue. This paper systematically explores the methodology of coreference resolution in IQA. Coreference information and discourse structures are labeled in IQA text and experiment on the IQA text to study the flexibility of the baseline system. Then it proposes several features about discourse structures in IQA. Experimental results show that this features are effective.

Key words: interactive question answering, discourse structure, coreference resolution