Argument Mining Workshop 2025

March 25th 2025 Happy to announce our Keynote Speaker: Andreas Vlachos from University of Cambridge.

January 28th 2025 12th Argument mining workshop will host two Shared Tasks

January 15th 2025 Check the Important Dates

January 15th 2025 The 1st Call for Papers is out!

December 6th 2024 Call for Shared Task is out

December 6th 2024 The official ArgMining 2025 website is launched.

Argument Mining (also known as “argumentation mining”) is an emerging research area within computational linguistics that started with focusing on automatically identifying and classifying argument elements, covering several text genres such as legal documents, news articles, online debates, scholarly data, and many more. In recent years, the field (broadly Computational Argumentation) has grown to explore argument quality and synthesis on many levels. The field offers practical uses such as argument-focused search and debating technologies, e.g., IBM Project Debater. The growing interest in computational argumentation has led to several tutorials at major NLP conferences.

Besides providing a forum to discuss and exchange cutting edge research in this field, a secondary goal of this year's edition will be to broaden the disciplinary scope of the workshop by inviting other disciplines (e.g., (computational) social and political science, psychology, humanities) as well as other subareas of NLP to actively participate in the workshop and further shaping the field of argument mining. In particular, we would like to create synergies between the fields of argument mining and natural language reasoning.

The workshop will be co-located with ACL 2025 and held in Vienna, Austria in a hybrid format.

Important Dates

Workshop: July 31st, 2025
Direct paper submission due (OpenReview): April 17th, 2025
Commitment deadline for ARR papers (OpenReview): May 21st, 2025
Notification of acceptance: May 28th, 2025
Camera-ready papers due: June 4th, 2025

All deadlines are 11.59 pm UTC -12h (“anywhere on Earth”).

Keynote Speaker

Andreas Vlachos Department of Computer Science and Technology, University of Cambridge

Title: TBA

About the Speaker: Andreas Vlachos is a professor of NLP and Machine Learning at the University of Cambridge. Among the many things he has worked on, we find constructiveness in argumentation, fact checking, media bias, dialogue modeling.

Submission Topics

The topics for submissions include but are not limited to:

Identification, Assessment, and Analysis of Arguments

Identification of argument components (e.g., premises and conclusions)
Structure analysis of arguments within and across documents
Relation Identification between arguments and counterarguments (e.g., support and attack)
Creation and evaluation of argument annotation schemes, relationships to linguistic and discourse annotations, (semi-) automatic argument annotation methods and tools, and creation of argumentation corpora
Assessment of arguments with respect to various properties (e.g., stance, clarity)

Generation of Arguments, Multi-modal and Multi-lingual Argument Mining

Automatic generation of arguments and their components
Consideration of discourse goals in argument generation
Argument mining and generation from multi-modal/multi-lingual data

Mining and Analysis of different Genres and Domains of Arguments

Argument mining in specific genres and domains (e.g., education, law, scientific writing)
Analysis of unique styles within genres (e.g., short informal text, highly structured writing)

Modelling, assessing, and critically reflecting on the argumentative reasoning capabilities of Large Language Models
Knowledge Integration, Information Retrieval, and Real-world Applications

Integration of commonsense and domain knowledge into argumentation models
Combination of information retrieval methods with argument mining
Real-world applications, including argument web search, opinion analysis and summarization, and misinformation detection

Interdisciplinary interfaces of Argument Mining
- Mining political discourse, by experts and laypeople
- Argument mining support for deliberation
- Persuasion and convincingess from a psychological perspective
- Subjectivity, disagreements and perspectivism in argumentation
Ethical Considerations and Future Reflections

Reflection on the ethical aspects and societal impact of argument mining methods
Reflection on the future of argument mining in light of the fast advancement of large language models (LLMs)

CALL FOR PAPERS

The Workshop on Argument Mining provides a regular forum for presenting and discussing cutting-edge research in argument mining (a.k.a argumentation mining) for academic and industry researchers. By continuing a series of eleven successful previous workshops, this edition will welcome the submission of long and short papers, as well as extended abstracts and PhD proposals. It will also feature a number of shared tasks shared tasks and a keynote talk.

Check DATES and TOPICS.

Submission Details

The organizing committee welcomes submitting long papers, short papers, extended abstracts and PhD proposals. Accepted papers will be presented via oral or poster presentations. Long and short papers will be included in the ACL proceedings as workshop papers. Extended abstracts and PhD proposals will be non-archival.

Long paper submissions must describe substantial, original, completed, and unpublished work. Wherever appropriate, concrete evaluation and analysis should be included. Long papers must be at most eight pages, including title, text, figures, and tables. An unlimited number of pages is allowed for references. Two additional pages are allowed for appendices, and an extra page is allowed in the final version to address reviewers’ comments.
Short paper submissions must describe original and unpublished work. Please note that a short paper is not a shortened long paper. Instead, short papers should have a point that can be made in a few pages, such as a small, focused contribution, a negative result, or an interesting application nugget. Short papers must be at most four pages, including title, text, figures, and tables. An unlimited number of pages is allowed for references. One additional page is allowed for the appendix, and an extra page is allowed in the final version to address reviewers’ comments.
Extended abstracts must be at most two pages including references describing ongoing projects, interesting pieces of data or results, or already published work.
PhD proposals must describe PhD projects being or to be developed within the broad field of natural language argumentation processing. PhD proposals must be at most four pages including the main research directions or challenges being investigated, the specific contributions made (on the research direction), and the directions for the remaining work. A dedicated poster session will be hosted, allowing students to get feedback and discuss their work with a broad and multidisciplinary community.

Multiple Submissions

ArgMining 2025 will not consider any paper under review in a journal or another conference or workshop at the time of submission, and submitted papers must not be submitted elsewhere during the review period. ArgMining 2025 will also accept submissions of ARR-reviewed papers, provided that the ARR reviews and meta-reviews are available by the ARR commitment deadline (May 21st). However, ArgMining 2025 will not accept direct submissions that are actively under review in ARR, or that overlap significantly (>25%) with such submissions.

Submission Format

All long, short, and demonstration submissions must follow the two-column ACL 2025 format. Authors are expected to use the LaTeX or Microsoft Word style template LaTeX or Microsoft Word style template. Submissions must conform to the official ACL style guidelines contained in these templates. Submissions must be electronic and in PDF format.

Submission Link

Authors have to fill in the submission form in the OpenReview system and upload a PDF of their paper here before April 17, 2025, 11:59 pm UTC-12h (anywhere on earth).

Double Blind Review

ArgMining 2025 will follow the ACL policies preserving the integrity of double-blind review for long and short paper submissions. Papers must not include authors' names and affiliations. Furthermore, self-references or links (such as GitHub) that reveal the author’s identity, e.g., “We previously showed (Smith, 1991) …” must be avoided. Instead, use citations such as “Smith previously showed (Smith, 1991) …” Papers that do not conform to these requirements will be rejected without review. Papers should not refer, for further detail, to documents that are not available to the reviewers. For example, do not omit or redact important citation information to preserve anonymity. Instead, use the third person or named reference to this work, as described above (“Smith showed” rather than “we showed”). Papers may be accompanied by a resource (software and/or data) described in the paper, but these resources should also be anonymized. Unlike long and short papers, demo descriptions will not be anonymous. Demo descriptions should include the authors’ names and affiliations, and self-references are allowed.
Anonimity Period (taken from the ACL call for papers in verbatim for the most part) We follow the ACL Policies for Review and Citation. Submissions must be anonymized, but there is no anonymity period or limitation on posting or discussing non-anonymous preprints while the work is under peer review.

Best Paper Award

In order to recognize significant advancements in argument mining science and technology, ArgMining 2025 will include the Best Paper award. All papers at the workshop are eligible for the best paper award, and a selection committee consisting of prominent researchers in the fields of interest will select the award recipients.

Shared Tasks

The 12th Argument Mining Workshop will be hosting two shared tasks.

1. Critical Questions Generation Shared Task

Organizers : Blanca Calvo Figueras,Rodrigo Agerri , HiTZ Basque Center for Language Technology - Ixa, University of the Basque Country UPV/EHU, Spain,
Elena Cabrio,Serena Villata, University of Côte d’Azur and member of the Inria-I3S research team Wimmics

Overview: In recent years, a growing concern within the educational community has been whether the widespread use of LLM-based chats could foster superficial learning habits and weaken students’ critical thinking skills. To counter this trend, in this task, we propose using LLMs to guide users towards asking critical questions. That is, questions that can uncover fallacious or poorly constructed arguments. In short: we want to foster critical thinking by developing a system that generates insightful critical questions when given argumentative texts.

All the details regarding the shared task can be found at the Critical Questions Generation Shared Task Website.

2. MM-ArgFallacy2025: Multimodal Argumentative Fallacy Detection and Classification on Political Debates

Organizers: Eleonora Mancini, Federico Ruggeri, Paolo Torroni, Language Technologies Lab, University of Bologna, Italy
Serena Villata, Inria-I3S WIMMICS Laboratoire I3S, CNRS, Sophia Antipolis, France

Overview: Argumentative fallacies, logical errors in reasoning, play a significant role in shaping debates, influencing opinions, and spreading misinformation. While traditional approaches to Argument Mining have focused on text-based analysis, paralinguistic features such as tone, pitch, and delivery can provide additional insights into fallacious reasoning. This shared task introduces the first Multimodal Argumentative Fallacy Detection and Classification challenge, leveraging both textual and audio data from political debates. Participants will address two sub-tasks: detecting fallacious argumentative sentences and classifying fallacies into predefined categories. The task is structured to evaluate performance across three settings: text-only, audio-only, and multimodal (text + audio). To support these goals, we provide datasets including MM-USED-fallacy, UKDebates, M-Arg, and MM-USED, fostering multi-task and multimodal learning. This task aims to advance research in multimodal argument mining, encouraging innovative methods that integrate diverse data modalities to better understand and detect logical fallacies in human discourse.

All the details regarding the shared task can be found at the MM-ArgFallacy2025: Multimodal Argumentative Fallacy Detection and Classification on Political Debates.