Identification of Caused Motion Construction

Proceedings of the Fourth Joint Conference on Lexical and Computational Semantics (*SEM 2015), pages 51–60, Denver, Colorado, June 4–5, 2015. Identification of Caused Motion Constructions Jena D. Hwang University of Colorado at Boulder Boulder, CO 80309 hwangd@colorado.edu Martha Palmer University of Colorado at Boulder Boulder, CO 80309 martha.palmer@colorado.edu Abstract This research describes the development of a supervised classifier of English Caused Mo- tion Constructions (CMCs) (e.g. The goalie kicked the ball into the field). Consistent iden- tification of CMCs is a necessary step to a cor- rect interpretation of semantics for sentences where the verb does not conform to the ex- pected semantics of the verb (e.g. The crowd laughed the clown off the stage). We ex- pand on a previous study on the classifica- tion CMCs (Hwang et al., 2010) to show that CMCs can be successfully identified in the corpus data. In this paper, we present the clas- sifier and the series of experiments carried out to improve its performance. 1 Introduction While natural language processing performance has been improved through the recognition that there is a relationship between the semantics of the verb and the syntactic context in which the verb is real- ized (Guildea and Palmer, 2002), sentences where the verb does not conform to the expected syntax- semantic patterning behavior remain problematic. 1. The goalie kicked the ball into the field. 2. The crowd laughed the clown off the stage. These sentences are semantically related – an en- tity causes a second entity to go along the path de- scribed by the prepositional phrase: in 1, the goalie causes the ball to go into the field, and in 2, the crowd causes the clown to go off the stage. While only the verb in the first sentence is gen- erally identified as a verb of motion that can ap- pear in a caused motion context, both are examples of caused motion constructions (CMCs) (Goldberg, 1995). The verb laugh of sentence 2 is normally considered an intransitive manner of speaking verb (e.g. The crowd laughed at the clown), but in this sentence, the verb is coerced into the caused motion interpretation and the semantics of the verb gives the manner in which the movement happened (e.g. the crowd caused the clown to move off the stage by means of laughing). The semantics parallel one another: both sentences have a causal argument re- sponsible for the event, an argument in motion, and a path that specifies the initial, middle, or final lo- cation, state or condition of the argument in motion (Hwang et al., 2013). Thus, if the semantic interpretation is strictly based on the expected semantics of the verb and its arguments, it fails to include the relevant infor- mation from the CMC. Accurate semantic role la- belling requires that NLP classifiers accurately iden- tify these coerced usages in data. In a previous study, we carried out prelim- inary work on the supervised identification of CMCs (Hwang et al., 2010). The pilot study was conducted in …

Content truncated to 500 words