Predicting Discourse Connectives for Implicit Discourse Relation Recognition
Existing works indicate that the absence of explicit discourse connectives makes it dif cult to recognize implicit discourse relations.In this paper we attempt to overcome this dif culty for implicit rela- tion recognition by automatically insert- ing discourse connectives between argu- ments with the use of a language model. Then we propose two algorithms to lever- age the information of these predicted connectives.One is to use these pre- dicted implicit connectives as additional features in a supervised model.The other is to perform implicit relation recognition based only on these predicted connectives. Results on Penn Discourse Treebank 2.0 show that predicted discourse connectives help implicit relation recognition and the rst algorithm can achieve an absolute av- erage f-score improvement of 3％over a state of the art baseline system.
Zhi-Min Zhou Yu Xu Zheng-Yu Niu Man Lan Jian Su Chew Lim Tan
East China Normal University Toshiba China R&D Center Institute for Infocomm Research National University of Singapore
国际会议
The 23rd International Conference on Computational Linguistics(第23届国际计算语言学大会)
北京
英文
1507-1514
2010-08-01（万方平台首次上网日期，不代表论文的发表时间）