Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing 2017
DOI: 10.18653/v1/d17-1261
|View full text |Cite
|
Sign up to set email alerts
|

Towards Debate Automation: a Recurrent Model for Predicting Debate Winners

Abstract: In this paper we introduce a practical first step towards the creation of an automated debate agent: a state-of-the-art recurrent predictive model for predicting debate winners. By having an accurate predictive model, we are able to objectively rate the quality of a statement made at a specific turn in a debate. The model is based on a recurrent neural network architecture with attention, which allows the model to effectively account for the entire debate when making its prediction. Our model achieves state-of… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
2

Citation Types

0
14
0
2

Year Published

2018
2018
2023
2023

Publication Types

Select...
8
1
1

Relationship

0
10

Authors

Journals

citations
Cited by 22 publications
(16 citation statements)
references
References 27 publications
0
14
0
2
Order By: Relevance
“…There is a rich body of prior work on classifying the outcome of a conversation after it has concluded, or classifying conversational events after they happened. Many examples exist, but some more closely related to our present work include identifying the winner of a debate (Zhang et al, 2016;Potash and Rumshisky, 2017;Wang et al, 2017), identifying successful negotiations (Curhan and Pentland, 2007;Cadilhac et al, 2013), as well as detecting whether deception (Girlea et al, 2016;Pérez-Rosas et al, 2016;Levitan et al, 2018) or disagreement (Galley et al, 2004;Abbott et al, 2011;Allen et al, 2014;Wang and Cardie, 2014;Rosenthal and McKeown, 2015) has occurred.…”
Section: Further Related Workmentioning
confidence: 88%
“…There is a rich body of prior work on classifying the outcome of a conversation after it has concluded, or classifying conversational events after they happened. Many examples exist, but some more closely related to our present work include identifying the winner of a debate (Zhang et al, 2016;Potash and Rumshisky, 2017;Wang et al, 2017), identifying successful negotiations (Curhan and Pentland, 2007;Cadilhac et al, 2013), as well as detecting whether deception (Girlea et al, 2016;Pérez-Rosas et al, 2016;Levitan et al, 2018) or disagreement (Galley et al, 2004;Abbott et al, 2011;Allen et al, 2014;Wang and Cardie, 2014;Rosenthal and McKeown, 2015) has occurred.…”
Section: Further Related Workmentioning
confidence: 88%
“…In contrast, our data set contains both same stance pairs, as well as cross stance pairs (i.e., one is supporting and the other is contesting the topic). Thus it is aligned with the above mentioned task, but in addition, with the task of choosing which side of the debate was more convincing (Potash and Rumshisky, 2017).…”
Section: Introductionmentioning
confidence: 82%
“…User factors are explored in previous papers Cardie, 2019a, 2018;Longpre et al, 2019), demonstrating the importance of characteristics and beliefs of the audience. Furthermore, Potash and Rumshisky (2017) proposed a recurrent neural network architecture with attention and annotated audience favorability to predict the winner of the debate. Villata et al (2018) and Benlamine et al (2017) studied the correlation of the engagement index in brain hemispheres with the persuasion strategies.…”
Section: Related Workmentioning
confidence: 99%