CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining

Debela Gemechu; Chris Reed

doi:10.18653/v1/2025.acl-long.969

CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining

Computing

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

30 Downloads (Pure)

Abstract

Argument Mining (AM) involves the automatic identification of argument structure in natural language. Traditional AM methods rely on micro-structural features derived from the internal properties of individual Argumentative Discourse Units (ADUs). However, argument structure is shaped by a macro-structure capturing the functional interdependence among ADUs. This macro-structure consists of segments, where each segment contains ADUs that fulfill specific roles to maintain coherence within the segment (**local coherence**) and across segments (**global coherence**). This paper presents an approach that models macro-structure, capturing both local and global coherence to identify argument structures. Experiments on heterogeneous datasets demonstrate superior performance in both in-dataset and cross-dataset evaluations. The cross-dataset evaluation shows that macro-structure enhances transferability to unseen datasets.

Original language	English
Title of host publication	Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics
Editors	Wanxiang Che, Joyce Nabende, Ekaterina Shutova, Mohammad Taher Pilehvar
Place of Publication	Kerrville, TX
Publisher	Association for Computational Linguistics
Pages	19731–19749
Number of pages	19
Volume	1 (Long Papers)
ISBN (Print)	9798891762510
DOIs	https://doi.org/10.18653/v1/2025.acl-long.969
Publication status	Published - Jul 2025
Event	The 63rd Annual Meeting of the Association for Computational Linguistics - Austria Center Vienna, Vienna, Austria Duration: 27 Jul 2025 → 1 Aug 2025 https://2025.aclweb.org/

Conference

Conference	The 63rd Annual Meeting of the Association for Computational Linguistics
Abbreviated title	ACL 2025
Country/Territory	Austria
City	Vienna
Period	27/07/25 → 1/08/25
Internet address	https://2025.aclweb.org/

Access to Document

10.18653/v1/2025.acl-long.969Licence: CC BY

Final Published VersionFinal published version, 1.39 MBLicence: CC BY

1 Finished
1 Active

Deliberation Laboratory (Joint with University of Gottingen, University of Maastricht, Warsaw University of Technology And University of Konstanz)
Reed, C. (Investigator)
1/07/21 → 31/12/26
Project: Research
AI for Citizen Intelligent Coaching against Disinformation (TITAN)
Lawrence, J. (Investigator) & Reed, C. (Investigator)
1/09/22 → 30/11/25
Project: Research

Cite this

Gemechu, D., & Reed, C. (2025). CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining. In W. Che, J. Nabende, E. Shutova, & M. T. Pilehvar (Eds.), Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Vol. 1 (Long Papers), pp. 19731–19749). Association for Computational Linguistics. https://doi.org/10.18653/v1/2025.acl-long.969

Gemechu, Debela ; Reed, Chris. / CU-MAM : Coherence-Driven Unified Macro-Structures for Argument Mining. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics. editor / Wanxiang Che ; Joyce Nabende ; Ekaterina Shutova ; Mohammad Taher Pilehvar. Vol. 1 (Long Papers) Kerrville, TX : Association for Computational Linguistics, 2025. pp. 19731–19749

@inproceedings{c2a18708b2934d35b3f8c9388b409ac7,

title = "CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining",

abstract = "Argument Mining (AM) involves the automatic identification of argument structure in natural language. Traditional AM methods rely on micro-structural features derived from the internal properties of individual Argumentative Discourse Units (ADUs). However, argument structure is shaped by a macro-structure capturing the functional interdependence among ADUs. This macro-structure consists of segments, where each segment contains ADUs that fulfill specific roles to maintain coherence within the segment (**local coherence**) and across segments (**global coherence**). This paper presents an approach that models macro-structure, capturing both local and global coherence to identify argument structures. Experiments on heterogeneous datasets demonstrate superior performance in both in-dataset and cross-dataset evaluations. The cross-dataset evaluation shows that macro-structure enhances transferability to unseen datasets.",

author = "Debela Gemechu and Chris Reed",

note = "Copyright: {\textcopyright} 2025 Association for Computational Linguistics ; The 63rd Annual Meeting of the Association for Computational Linguistics, ACL 2025 ; Conference date: 27-07-2025 Through 01-08-2025",

year = "2025",

month = jul,

doi = "10.18653/v1/2025.acl-long.969",

language = "English",

isbn = "9798891762510",

volume = "1 (Long Papers)",

pages = "19731{\textendash}19749",

editor = "Che, \{Wanxiang \} and Nabende, \{Joyce \} and Shutova, \{Ekaterina \} and Pilehvar, \{Mohammad Taher\}",

booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics",

publisher = "Association for Computational Linguistics",

url = "https://2025.aclweb.org/",

}

Gemechu, D & Reed, C 2025, CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining. in W Che, J Nabende, E Shutova & MT Pilehvar (eds), Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics. vol. 1 (Long Papers), Association for Computational Linguistics, Kerrville, TX, pp. 19731–19749, The 63rd Annual Meeting of the Association for Computational Linguistics, Vienna, Austria, 27/07/25. https://doi.org/10.18653/v1/2025.acl-long.969

CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining. / Gemechu, Debela ; Reed, Chris.
Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics. ed. / Wanxiang Che; Joyce Nabende; Ekaterina Shutova; Mohammad Taher Pilehvar. Vol. 1 (Long Papers) Kerrville, TX: Association for Computational Linguistics, 2025. p. 19731–19749.

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

TY - GEN

T1 - CU-MAM

T2 - The 63rd Annual Meeting of the Association for Computational Linguistics

AU - Gemechu, Debela

AU - Reed, Chris

PY - 2025/7

Y1 - 2025/7

N2 - Argument Mining (AM) involves the automatic identification of argument structure in natural language. Traditional AM methods rely on micro-structural features derived from the internal properties of individual Argumentative Discourse Units (ADUs). However, argument structure is shaped by a macro-structure capturing the functional interdependence among ADUs. This macro-structure consists of segments, where each segment contains ADUs that fulfill specific roles to maintain coherence within the segment (**local coherence**) and across segments (**global coherence**). This paper presents an approach that models macro-structure, capturing both local and global coherence to identify argument structures. Experiments on heterogeneous datasets demonstrate superior performance in both in-dataset and cross-dataset evaluations. The cross-dataset evaluation shows that macro-structure enhances transferability to unseen datasets.

AB - Argument Mining (AM) involves the automatic identification of argument structure in natural language. Traditional AM methods rely on micro-structural features derived from the internal properties of individual Argumentative Discourse Units (ADUs). However, argument structure is shaped by a macro-structure capturing the functional interdependence among ADUs. This macro-structure consists of segments, where each segment contains ADUs that fulfill specific roles to maintain coherence within the segment (**local coherence**) and across segments (**global coherence**). This paper presents an approach that models macro-structure, capturing both local and global coherence to identify argument structures. Experiments on heterogeneous datasets demonstrate superior performance in both in-dataset and cross-dataset evaluations. The cross-dataset evaluation shows that macro-structure enhances transferability to unseen datasets.

U2 - 10.18653/v1/2025.acl-long.969

DO - 10.18653/v1/2025.acl-long.969

M3 - Conference contribution

SN - 9798891762510

VL - 1 (Long Papers)

SP - 19731

EP - 19749

BT - Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics

A2 - Che, Wanxiang

A2 - Nabende, Joyce

A2 - Shutova, Ekaterina

A2 - Pilehvar, Mohammad Taher

PB - Association for Computational Linguistics

CY - Kerrville, TX

Y2 - 27 July 2025 through 1 August 2025

ER -

Gemechu D , Reed C. CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining. In Che W, Nabende J, Shutova E, Pilehvar MT, editors, Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics. Vol. 1 (Long Papers). Kerrville, TX: Association for Computational Linguistics. 2025. p. 19731–19749 doi: 10.18653/v1/2025.acl-long.969

CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining

Abstract

Conference

Access to Document

Fingerprint

Projects

Deliberation Laboratory (Joint with University of Gottingen, University of Maastricht, Warsaw University of Technology And University of Konstanz)

AI for Citizen Intelligent Coaching against Disinformation (TITAN)

Cite this