Corpus-based Open-Domain Event Type Induction

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 5427–5440 November 7–11, 2021. c⃝2021 Association for Computational Linguistics 5427 Corpus-based Open-Domain Event Type Induction Jiaming Shen, Yunyi Zhang, Heng Ji, Jiawei Han Department of Computer Science, University of Illinois Urbana-Champaign, IL, USA {js2, yzhan238, hengji, hanj}@illinois.edu Abstract Traditional event extraction methods require predefined event types and their corresponding annotations to learn event extractors. These prerequisites are often hard to be satisfied in real-world applications. This work presents a corpus-based open-domain event type induc- tion method that automatically discovers a set of event types from a given corpus. As events of the same type could be expressed in multi- ple ways, we propose to represent each event type as a cluster of ⟨predicate sense, object head⟩ pairs. Specifically, our method (1) se- lects salient predicates and object heads, (2) disambiguates predicate senses using only a verb sense dictionary, and (3) obtains event types by jointly embedding and clustering ⟨predicate sense, object head⟩ pairs in a latent spherical space. Our experiments, on three datasets from different domains, show our method can discover salient and high-quality event types, according to both automatic and human evaluations1. 1 Introduction One step towards converting massive unstructured text into structured, machine-readable representa- tions is event extraction—the identification and typ- ing of event triggers and arguments in text. Most event extraction methods (Ahn, 2006; Ji and Grish- man, 2008; Du and Cardie, 2020; Li et al., 2021) assume a set of predefined event types and their corresponding annotations are curated by human experts. This annotation process is expensive and time-consuming. Besides, those manually-defined event types often fail to generalize to new do- mains. For example, the widely used ACE 2005 event schemas2 do not contain any event type 1The programs, data and resources are publicly avail- able for research purpose at https://github.com/ mickeystroller/ETypeClus. 2https://www.ldc.upenn.edu/ collaborations/past-projects/ace detain_1 people arrest_1 people arrest_2 spread stop_1 transmission stop_1 planning “Arrest-Jail” “Stop-Spread” “Stop-Plan” Sentences Hundreds of people are detained for distributing purported false information online. Researchers say that vaccinating 46 percent of Haitians could arrest the cholera spread. The Zimbabwe CTU said 69 people were arrested during Wednesday's demonstrations. More censorship of social media posts are enforced to stop protest planning online. Collective efforts are needed by all nations to stop the COVID-19 transmission. ID S5 S4 S3 S2 S1 Figure 1: Motivating example sentences and induced event types. Predicates are in bold. Objects are under- lined and object heads are in italics. Colors indicate event types. The suffix number followed by each pred- icate verb lemma indicates the predicate verb sense. about Transmit Virus or Treat Disease and thus cannot be readily applied to extract pan- demic events. To automatically induce event schemas from raw text, researchers have studied ad-hoc clustering- based algorithms (Sekine, 2006; Chambers and Jurafsky, 2011) and probabilistic generative meth- ods (Chambers, 2013; Cheung et al., 2013; Nguyen et al., 2015) to discover a set of event types and argument roles. These methods typically utilize …

Content truncated to 500 words