Multi-Relational Script Learning for Discourse Relations

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 4214–4226 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 4214 Multi-Relational Script Learning for Discourse Relations I-Ta Lee and Dan Goldwasser Department of Computer Science, Purdue University {lee2226, dgoldwas}@purdue.edu Abstract Modeling script knowledge can be useful for a wide range of NLP tasks. Current statistical script learning approaches embed the events, such that their relationships are indicated by their similarity in the embedding. While in- tuitive, these approaches fall short of repre- senting nuanced relations, needed for down- stream tasks. In this paper, we suggest to view learning event embedding as a multi-relational problem, which allows us to capture different aspects of event pairs. We model a rich set of event relations, such as Cause and Contrast, derived from the Penn Discourse Tree Bank. We evaluate our model on three types of tasks, the popular Mutli-Choice Narrative Cloze and its variants, several multi-relational prediction tasks, and a related downstream task—implicit discourse sense classification. 1 Introduction Representing world knowledge that can be used for commonsense reasoning is a long-standing AI goal. Scripts (Schank and Abelson, 1977) are structured knowledge representations captur- ing the relationships between prototypical event sequences and their participants in a given sce- nario. For example, given the event “John shot Jim with a gun”, we can infer that “he got arrested by police” is more probable than “he fell asleep”. In recent years, the problem of extracting script knowledge from text has attracted significant at- tention. Early works (Chambers and Jurafsky, 2008) focused on symbolic event representations and used Pointwise Mutual Information (PMI) be- tween events to capture their relationships. Recent works (Pichotta and Mooney, 2016a; Granroth- Wilding and Clark, 2016; Wang et al., 2017; Lee and Goldwasser, 2018; Li et al., 2018) represent events using dense vectors, based on event co- occurrence, and use vector similarity over their embeddings to measure their relationship. Our main observation in this paper is that while models for learning script knowledge improved significantly over the last decade, these models can essentially represent only a single event rela- tionship, co-occurrence. That is, events appearing in similar contexts tend to have similar represen- tations. Although this idea works well for a lot of NLP tasks, it is too coarse for modeling com- monsense, which should account for fine-grained relationships. To better understand this, consider the example described in Figure 1. Given the first event, corresponding to the sentence “Jenny went to her favorite restaurant.”, called Step 1, any of the following events in Step 2 would be highly re- lated, and thus similar, to the input event. That is, “It was raining outside” and “She was very hun- gry” are both possible NEXT events. Using event similarity alone is too coarse to support many rel- evant inferences. However, if the relation between the events is given, more clues can be applied to support reliable inferences. In Figure 1, given Step 2 is a Reason to …

Content truncated to 500 words