To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.ipm.2013.10.003 Byline: Pei-Feng Li, Qiao-Ming Zhu, Guo-Dong Zhou Abstract: acents We employ the compositional semantics to infer unknown triggers. acents We impose the discourse consistency to recover missing trigger mentions. acents The morphological structures are better to represent the compositional semantics. acents The head morpheme is the key clue to infer unknown triggers. Article History: Received 23 October 2012; Revised 11 October 2013; Accepted 18 October 2013 Article Note: (footnote) [star] This paper is a much extended version of Li et al. (2012) at EMNLP-CoNLL 2012.