Combining Extraction and Generation for Constructing Belief-Consequence Causal Links

Maria Alexeeva; Allegra A. Beal Cohen; Mihai Surdeanu

Combining Extraction and Generation for Constructing Belief-Consequence Causal Links

Maria Alexeeva
, Allegra A. Beal Cohen
, Mihai Surdeanu

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

In this paper, we introduce and justify a new task—causal link extraction based on beliefs—and do a qualitative analysis of the ability of a large language model—InstructGPT-3—to generate implicit consequences of beliefs. With the language model-generated consequences being promising, but not consistent, we propose directions of future work, including data collection, explicit consequence extraction using rule-based and language modeling-based approaches, and using explicitly stated consequences of beliefs to fine-tune or prompt the language model to produce outputs suitable for the task.

Original language	English (US)
Title of host publication	Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop
Editors	Shabnam Tafreshi, Joao Sedoc, Anna Rogers, Aleksandr Drozd, Anna Rumshisky, Arjun Reddy Akula
Publisher	Association for Computational Linguistics (ACL)
Pages	159-164
Number of pages	6
ISBN (Electronic)	9781955917407
State	Published - 2022
Event	3rd Workshop on Insights from Negative Results in NLP, Insights 2022 - Dublin, Ireland Duration: May 26 2022 → …

Publication series

Name	Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop

Conference

Conference	3rd Workshop on Insights from Negative Results in NLP, Insights 2022
Country/Territory	Ireland
City	Dublin
Period	5/26/22 → …

ASJC Scopus subject areas

Language and Linguistics
Computational Theory and Mathematics
Computer Science Applications
Software

Cite this

Alexeeva, M., Beal Cohen, A. A., & Surdeanu, M. (2022). Combining Extraction and Generation for Constructing Belief-Consequence Causal Links. In S. Tafreshi, J. Sedoc, A. Rogers, A. Drozd, A. Rumshisky, & A. R. Akula (Eds.), Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop (pp. 159-164). (Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop). Association for Computational Linguistics (ACL).

Combining Extraction and Generation for Constructing Belief-Consequence Causal Links. / Alexeeva, Maria; Beal Cohen, Allegra A.; Surdeanu, Mihai.
Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop. ed. / Shabnam Tafreshi; Joao Sedoc; Anna Rogers; Aleksandr Drozd; Anna Rumshisky; Arjun Reddy Akula. Association for Computational Linguistics (ACL), 2022. p. 159-164 (Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Alexeeva, M, Beal Cohen, AA & Surdeanu, M 2022, Combining Extraction and Generation for Constructing Belief-Consequence Causal Links. in S Tafreshi, J Sedoc, A Rogers, A Drozd, A Rumshisky & AR Akula (eds), Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop. Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop, Association for Computational Linguistics (ACL), pp. 159-164, 3rd Workshop on Insights from Negative Results in NLP, Insights 2022, Dublin, Ireland, 5/26/22.

Alexeeva M, Beal Cohen AA, Surdeanu M. Combining Extraction and Generation for Constructing Belief-Consequence Causal Links. In Tafreshi S, Sedoc J, Rogers A, Drozd A, Rumshisky A, Akula AR, editors, Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop. Association for Computational Linguistics (ACL). 2022. p. 159-164. (Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop).

Alexeeva, Maria ; Beal Cohen, Allegra A. ; Surdeanu, Mihai. / Combining Extraction and Generation for Constructing Belief-Consequence Causal Links. Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop. editor / Shabnam Tafreshi ; Joao Sedoc ; Anna Rogers ; Aleksandr Drozd ; Anna Rumshisky ; Arjun Reddy Akula. Association for Computational Linguistics (ACL), 2022. pp. 159-164 (Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop).

@inproceedings{416ca8d53c6246be9811b51902b1f60d,

title = "Combining Extraction and Generation for Constructing Belief-Consequence Causal Links",

abstract = "In this paper, we introduce and justify a new task—causal link extraction based on beliefs—and do a qualitative analysis of the ability of a large language model—InstructGPT-3—to generate implicit consequences of beliefs. With the language model-generated consequences being promising, but not consistent, we propose directions of future work, including data collection, explicit consequence extraction using rule-based and language modeling-based approaches, and using explicitly stated consequences of beliefs to fine-tune or prompt the language model to produce outputs suitable for the task.",

author = "Maria Alexeeva and \{Beal Cohen\}, \{Allegra A.\} and Mihai Surdeanu",

note = "Publisher Copyright: {\textcopyright} 2022 Association for Computational Linguistics.; 3rd Workshop on Insights from Negative Results in NLP, Insights 2022 ; Conference date: 26-05-2022",

year = "2022",

language = "English (US)",

series = "Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop",

publisher = "Association for Computational Linguistics (ACL)",

pages = "159--164",

editor = "Shabnam Tafreshi and Joao Sedoc and Anna Rogers and Aleksandr Drozd and Anna Rumshisky and Akula, \{Arjun Reddy\}",

booktitle = "Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop",

}

TY - GEN

T1 - Combining Extraction and Generation for Constructing Belief-Consequence Causal Links

AU - Alexeeva, Maria

AU - Beal Cohen, Allegra A.

AU - Surdeanu, Mihai

PY - 2022

Y1 - 2022

N2 - In this paper, we introduce and justify a new task—causal link extraction based on beliefs—and do a qualitative analysis of the ability of a large language model—InstructGPT-3—to generate implicit consequences of beliefs. With the language model-generated consequences being promising, but not consistent, we propose directions of future work, including data collection, explicit consequence extraction using rule-based and language modeling-based approaches, and using explicitly stated consequences of beliefs to fine-tune or prompt the language model to produce outputs suitable for the task.

AB - In this paper, we introduce and justify a new task—causal link extraction based on beliefs—and do a qualitative analysis of the ability of a large language model—InstructGPT-3—to generate implicit consequences of beliefs. With the language model-generated consequences being promising, but not consistent, we propose directions of future work, including data collection, explicit consequence extraction using rule-based and language modeling-based approaches, and using explicitly stated consequences of beliefs to fine-tune or prompt the language model to produce outputs suitable for the task.

UR - https://www.scopus.com/pages/publications/85137478501

UR - https://www.scopus.com/pages/publications/85137478501#tab=citedBy

M3 - Conference contribution

AN - SCOPUS:85137478501

T3 - Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop

SP - 159

EP - 164

BT - Insights 2022 - 3rd Workshop on Insights from Negative Results in NLP, Proceedings of the Workshop

A2 - Tafreshi, Shabnam

A2 - Sedoc, Joao

A2 - Rogers, Anna

A2 - Drozd, Aleksandr

A2 - Rumshisky, Anna

A2 - Akula, Arjun Reddy

PB - Association for Computational Linguistics (ACL)

T2 - 3rd Workshop on Insights from Negative Results in NLP, Insights 2022

Y2 - 26 May 2022

ER -

Combining Extraction and Generation for Constructing Belief-Consequence Causal Links

Abstract

Publication series

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this