Salience-Aware Event Chain Modeling for Narrative Understanding

Abstract

Storytelling, whether via fables, news reports, documentaries, or memoirs, can be thought of as the communication of interesting and related events that, taken together, form a concrete process. It is desirable to extract the event chains that represent such processes. However, this extraction remains a challenging problem. We posit that this is due to the nature of the texts from which chains are discovered.Natural language text interleaves a narrative of concrete, salient events with background information, contextualization, opinion, and other elements that are important for a variety of necessary discourse and pragmatics acts but are not part of the principal chain of events being communicated. We introduce methods for extracting this principal chain from natural language text, by filtering away non-salient events and supportive sentences. We demonstrate the effectiveness of our methods at isolating critical event chains by comparing their effect on downstream tasks. We show that by pre-training large language models on our extracted chains, we obtain improvements in two tasks that benefit from a clear understanding of event chains: narrative prediction and event-based temporal question answering. The demonstrated improvements and ablative studies confirm that our extraction method isolates critical event chains. 1

Are NLP Models Good at Tracing Thoughts: An Overview of Narrative Understanding

“…Similarly, the ROC story cloze task (Mostafazadeh et al, 2016), addressed by Cai et al (2017), involves choosing the most plausible ending. There are various approaches developed for story ending prediction, such as the incorporation of commonsense knowledge (Li et al, 2018b), utilization of skip-thought embeddings (Srinivasan et al, 2018), entity-driven recurrent networks (Henaff et al, 2017;Liu et al, 2018), scene structure (Tian et al, 2020), centrality or salience of events (Zhang et al, 2021), and contextualized narrative event representation (Wilner et al, 2021), respectively. Simple and wellestablished, the Story Cloze Test does not cover the core aspects of narrative structure, though.…”

Identifying Conspiracy Theories News based on Event Relation Graph

“…Next, we train four event relation extractors. Following the previous work (Zhang et al, 2021;Yao et al, 2020), we form the training event pairs in the natural textual order, where the former event in the pair is the precedent event mentioned in text. Regarding temporal relations 2 , we process the before annotation as follows: keep the label as before if the annotated event pair aligns with the natural textual order, or assign after label to the reverse pair if not.…”

“…One observation we have for identifying conspiracy theories is that a conspiracy story can be made up by mixing unrelated events together. The concept of "event" refers to a specific occurrence or action that happens around us, and is essential in story telling and narrative understanding (Zhang et al, 2021). Conspiracy theories, however, usually put typically unrelated events together to fabricate a false story.…”

NECE: Narrative Event Chain Extraction Toolkit

“…Due to limitations of current method, we only offer an imperfect gender categorization as female, male, group/non-binary, or unknown (See Limitation 5) #7. Salient events are defined as events that are relevant, or essential, to the main narrative of the story as defined by (Zhang et al, 2021). To find the salient events, we trained an inverse document frequency(idf) dictionary on a fairy-tale corpus (Xu et al, 2022) to calculate a tf-idf based salience score.…”