DEMN: Distilled-Exposition Enhanced Matching Network for Story Comprehension

DEMN: Distilled-Exposition Enhanced Matching Network for Story Comprehension Chunhua Liu1 Haiou Zhang1 Shan Jiang1 Dong Yu1,2 � 1 Beijing Language and Culture University 2 Beijing Advanced Innovation for Language Resources of BLCU {chunhualiu596, jiangshan727}@gmail.com {hozhangel, yudong blcu}@126.com Abstract This paper proposes a Distilled-Exposition Enhanced Matching Network (DEMN) for story-cloze test, which is still a challenging task in story comprehension. We divide a complete story into three narrative segments: an exposition, a climax, and an ending. The model consists of three modules: input mod- ule, matching module, and distillation mod- ule. The input module provides semantic rep- resentations for the three segments and then feeds them into the other two modules. The matching module collects interaction features between the ending and the climax. The distil- lation module distills the crucial semantic in- formation in the exposition and infuses it into the matching module in two different ways. We evaluate our single and ensemble model on ROCStories Corpus (Mostafazadeh et al., 2016), achieving an accuracy of 80.1% and 81.2% on the test set respectively. The exper- imental results demonstrate that our DEMN model achieves a state-of-the-art performance. 1 Introduction Story comprehension is a fascinating task in natural language understanding with a long history (Jones, 1974; Turner, 1994). The difficulty of this task arises from the necessity of commonsense knowl- edge, cross-sentence reasoning, and causal reason- ing between events. The recently emerged story- cloze test (Mostafazadeh et al., 2016) focuses on commonsense story comprehension, which aims at choosing the most plausible ending from two op- tions for a four-sentence story (also called plot). (a) option & plot attention (b) option & climax (c) option & climax atten- tion (d) expostion enhanced op- tion & climax attention Figure 1: Four strategies of modeling a story Recently, methods based on linear classifier with handcrafted features, as well as neural network (NN), have been proposed for the story-cloze test. Handcrafted features based methods (Chaturvedi et al., 2017; Mostafazadeh et al., 2016; Schwartz et al., 2017) extract commonsense knowledge like events sequence and sentiments trajectory by external tools to help the story understanding. first Among NN based methods, the val-LS-skip model (Srinivasan et al., 2018) represents the last sentence in the plot and the option ending with skip-embeddings, and then processes them with a simple feed-forward neural network. Their experiments show that representing the whole four-sentence plot with the ending per- forms worse than only using the fourth sentence with the ending. PACLIC 32 385 32nd Pacific Asia Conference on Language, Information and Computation Hong Kong, 1-3 December 2018 Copyright 2018 by the authors However, intuitively speaking, a coherent ending should be related to the whole plot, instead of just the fourth sentence. This intuition is opposite to the conclusion of Srinivasan et al. (2018). To ex- plore whether the content in a plot, except for only the fourth sentence, can assist in choosing a correct ending, we observed a large number of stories and discovered two phenomena: (1) The ending is usu- ally directly affected by …

Content truncated to 500 words