Event Factuality Detection in Discourse

Huang, Rongtao; Zou, Bowei; Wang, Hongling; Li, Peifeng; Zhou, Guodong

doi:10.1007/978-3-030-32236-6_36

Cited by 9 publications

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Event factuality identification (EFI) is an important task in information extraction, which can be helpful for event extraction [6,24,43,44] and knowledge acquisition [42,21,22]. Pioneering studies explore the sentence-level EFI task, which have explored rule-based methods [26,39], machine learning methods [30,40,20,33] and neural networks [38,34,14,29]. However, an event can have conflicting factuality in different sentences, usually leading to confusing factuality results in applications.…”

Section: Related Workmentioning

confidence: 99%

Uncertain Relational Hypergraph Attention Networks for Document-Level Event Factuality Identification

Sheng,

Cong,

Cao

et al. 2023

Frontiers in Artificial Intelligence and Applications

View full text Add to dashboard Cite

Document-level event factuality identification (DocEFI) is an important task in event knowledge acquisition, which aims to detect whether an event actually occurs or not from the perspective of the document. Unlike the sentence-level task, a document can have multiple sentences with different event factualities, leading to event factuality conflicts in DocEFI. Existing studies attempt to aggregate local event factuality by exploiting document structures, but they mostly consider textual components in the document separately, degrading complicated correlations therein. To address the above issues, this paper proposes a novel approach, namely UR-HAT, to improve DocEFI with uncertain relational hypergraph attention networks. Particularly, we reframe a document graph as a hypergraph, and establish beneficial n-ary correlations among textual nodes with relational hyperedges, which helps to globally consider local factuality features to resolve event factuality conflicts. To better discern the importance of event factuality features, we further represent textual nodes with uncertain Gaussian distributions, and propose novel uncertain relational hypergraph attention networks to refine textual nodes with the document hypergraph. In addition, we select factuality-related keywords as nodes to enrich event factuality features. Experimental results demonstrate the effectiveness of our proposed method, and outperforms previous methods on two widely used benchmark datasets.

show abstract

Section: Related Workmentioning

confidence: 99%

Uncertain Relational Hypergraph Attention Networks for Document-Level Event Factuality Identification

Sheng,

Cong,

Cao

et al. 2023

Frontiers in Artificial Intelligence and Applications

View full text Add to dashboard Cite

show abstract

“…Regarding the automation of the annotation process, the systems currently available follow two different approaches: those using machine-learning techniques and those based, at least partially, on linguistic information. Among the former, Mullick et al (2019) present the development of a deep neural network based on the 'Factuality Judgment Model', while Huang et al (2019) use 'Bi-directional Long Short-Term Memory' (BiLSTM), that is, neural networks to learn contextual information about the event in sentences. The latter consider that annotating factuality at sentence level provides an incomplete picture and their unit of analysis is the document.…”

Section: Various Authors Have Drawn Onmentioning

confidence: 99%

The TAGFACT annotator and editor: A versatile tool

Fernández-Montraveta

Vázquez

Curell

et al. 2020

RiCL

View full text Add to dashboard Cite

The multifunctional tool this paper presents has been developed within the TAGFACT project, a project that aims to automate the annotation of factuality –understood as the degree of commitment with which the writer presents situations– in Spanish journalistic texts. In what follows, the tool, which allows the compilation of the texts and the manual annotation of predicates, is described. The corpus created using it has been extracted in groups of three pieces of news covering the same event from newspapers with different ideologies (left wing, right wing and centrist). It is made up of 176 different pieces of news, containing 1,359 sentences and 46,947 words. The tool has been used so far to manually annotate a section of the ‘Gold Standard’ (approximately 10,000 words). It has proved to be versatile in that it allows for both the creation and management of corpora and corpus annotation, using any tags the user wants depending on the purpose of each corpus.

show abstract

“…Given the context of a speciic social situation, humans can easily grasp the development line to predict subsequent events. However, representing such knowledge as a machine-readable format is still challenging in the artiicial intelligence ield, the success of which can support a series of downstream applications, e.g., question answering [25], discourse understanding [18] and information extraction [27], etc. Roger and Robert [39] pioneered to introduce the notion of script (a.k.a illmorean frames) to deine this schematic knowledge structure.…”

Section: Introductionmentioning

confidence: 99%

Sequence-aware Knowledge Distillation for a Lightweight Event Representation

Cai

Ling

Chen

2022

ACM Trans. Inf. Syst.

View full text Add to dashboard Cite

Event representation targets to model the event-reasoning process as a machine-readable format. Previous studies on event representation mostly concentrate on a sole modeling perspective and have not well investigated the scenario-level knowledge, which can cause information loss. To cope with this dilemma, we propose a unified fine-tuning architecture-based approach ( UniFA-S ) that integrates all levels of trainings, including the scenario-level knowledge. However, another challenge for existing models is the ever-increasing computation overheads, restricting the deployment ability on limited resources devices. Hence, in this paper, we aim to compress the cumbersome model UniFA-S into a lighter and easy-to-deploy one without much performance damage. To this end, we propose a sequence-aware knowledge distillation model (SaKD) that employs a dynamic self-distillation on the decouple-compress-couple framework for compressing UniFA-S , which cannot only realize the model compression but also retain the integrity of individual components. We also design two fitting strategies to address the less-supervised issue at the distillation stage. Comprehensive experiments on representation-and-inference ability-based tasks validate the effectiveness of SaKD. Compared to UniFA-S , SaKD realizes a more portable event representation model at the cost of only 1.0% performance drop in terms of accuracy or Spearman’s correlation that is far less than other knowledge distillation models.

show abstract

Event Factuality Detection in Discourse

Cited by 9 publications

References 9 publications

Uncertain Relational Hypergraph Attention Networks for Document-Level Event Factuality Identification

Uncertain Relational Hypergraph Attention Networks for Document-Level Event Factuality Identification

The TAGFACT annotator and editor: A versatile tool

Sequence-aware Knowledge Distillation for a Lightweight Event Representation

Contact Info

Product

Resources

About