Temporal Reasoning on Implicit Events from Distant Supervision

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1361–1371 June 6–11, 2021. ©2021 Association for Computational Linguistics 1361 Temporal Reasoning on Implicit Events from Distant Supervision Ben Zhou1,2 Kyle Richardson1 Qiang Ning3 Tushar Khot1 Ashish Sabharwal1 Dan Roth2 ∗ 1Allen Institute for AI 2University of Pennsylvania 3Amazon {kyler,tushark,ashishs}@allenai.org {xyzhou,danroth}@cis.upenn.edu qning@amazon.com Abstract We propose TRACIE, a novel temporal rea- soning dataset that evaluates the degree to which systems understand implicit events— events that are not mentioned explicitly in nat- ural language text but can be inferred from it. This introduces a new challenge in tem- poral reasoning research, where prior work has focused on explicitly mentioned events. Human readers can infer implicit events via commonsense reasoning, resulting in a more comprehensive understanding of the situation and, consequently, better reasoning about time. We find, however, that state-of-the-art mod- els struggle when predicting temporal relation- ships between implicit and explicit events. To address this, we propose a neuro-symbolic temporal reasoning model, SYMTIME, which exploits distant supervision signals from large- scale text and uses temporal rules to combine start times and durations to infer end times. SYMTIME outperforms strong baseline sys- tems on TRACIE by 5%, and by 11% in a zero prior knowledge training setting. Our ap- proach also generalizes to other temporal rea- soning tasks, as evidenced by a gain of 1%-9% on MATRES, an explicit event benchmark. 1 Introduction Understanding temporal relations between events in narrative text is a crucial part of text understand- ing. When reading a story, a human can construct a latent timeline about events’ start and end times, similar to the one shown in Fig. 1 about an auto- mobile accident. This timeline not only contains the placements of explicitly mentioned events (e.g., ride a bicycle), but also accounts for implicit events (e.g., Farrah was distracted so she looked away). Such a latent timeline explains the dynamics be- tween events; for example, the possible chain of events between ride and recovered in this context ∗Most of the work was done when the third author was employed at the Allen Institute for AI and the first author was an intern there. Farrah was driving home from school. A person was riding a bicycle in front of her. Farrah looked away for a second. She didn't notice that he stopped. She tried to brake but it was too late. The person recovered soon. Context Story Latent Timeline ride stopped get hit injured A person drive hit try regret get home distracted explicit events implicit events not-inferrable distracted starts before try ✅ entailment distracted ends after try ❌ contradiction recovered Tracie Instance Farrah …. many others starts starts look Figure 1: A story, its latent timeline, and example TRA- CIE instances from it. For simplicity, events are short- ened to single verbs and the timeline is exaggerated. contains get hit and injured. The ability to construct such a timeline is essential for understanding the causal dynamics of a …

Content truncated to 500 words