Centre for Argument Technology

Kamila Górska and John Lawrence are at #ECAI2025 presenting their work on Argumentative Strategies and Forecasting Success, and DRACS: Diachronic Representation of Argument Construction Styles.

This week, we discussed the paper "ImpScore: A Learnable Metric For Quantifying The Implicitness Level of Sentences" by Yuxin Wang et al., presenting an exciting method to move beyond semantic representations into indirect meaning. proceedings.iclr.cc/paper_files/...

This is the first study to systematically connect fine-grained argumentative structure with forecasting success, highlighting the role of structured reasoning in predictive accuracy.
📄 Read the full paper at: tinyurl.com/argstrategies

🧠 Forecasts that show balanced argumentative coverage, addressing multiple possible hypotheses rather than focusing on a single one, are associated with higher accuracy.
Graph-based models also outperform text-only baselines.

🔍 We model forecasters’ reasoning as argument graphs capturing support, conflict, and balance between propositions.
These graphs are enriched with text embeddings and analysed using a Graph Neural Network (GNN) to link structure and outcome.

📈 Accurate forecasting depends not only on data quality but also on the structure of reasoning behind predictions.
We examine how the organisation of arguments (how forecasters present, balance, and connect claims) relates to forecasting accuracy.

🎉 We are excited to present 'Argumentative Strategies and Forecasting Success’ at ECAI 2025! It investigates the role reasoning plays in predicting future events. If you are interested, read more details below! 👇 #ECAI2025 #Argumentation #Forecasting

📢 Our PhD student Kamila will present her paper "Argumentative Strategies and Forecasting Success" on Monday 27th October at ECAI 2025!
⚡Lighning Talk: 12:49-12:51
📄Poster Session: 13:00-14:00 (board 1-29)
See you all in Bologna!

📢 We are looking forward to presenting our paper "DRACS: Diachronic
Representation of Argument Construction Styles" by Yevhen Kostiuk, Chris Reed, John Lawrence on Monday 27th October at ECAI 2025. #ECAI2025 #Argumentation
⚡ Lighning Talk: 12:57-12:59
📄 Poster Session: 13:00-14:00 (board 2-157)

At this week’s reading group, we discussed the paper "On the Controllability of Large Language Models for Dialogue Interaction" by Nicolas Wagner and Stefan Ultes. aclanthology.org/2024.sigdial...

Today in our reading group we discussed the preprint, "The Thin Line Between Comprehension and Persuasion in LLMs", by Adrian de Wynter and Tangming Yuan, which annotates and evaluates debates with humans and LLMs.

📢 🎉 We're very happy to announce even more success for ARG-tech's researchers with a paper being accepted to the findings of #EMNLP2025! Congratulations to all authors! We're excited to see many of you in Suzhou, China this November 🇨🇳

📣 🎉 We are excited to announce that two papers authored by ARG-tech members have been accepted to #ECAI2025! Congratulations to our amazing reserachers for this great success.
We're looking forward to seeing many of you in Bologna in October 🇮🇹

In this week's reading group, we discussed the paper "Can Large Language Models Understand Argument Schemes?" recently published at the Findings of the ACL 2025 by @elfiabv.bsky.social, Oana Cocarascu, and Sanjay Modgil.
aclanthology.org/2025.finding...

Chris led this week's reading group discussion on a detailed study about annotation error detection (aclanthology.org/2023.cl-1.4). We focused on exploring how well these techniques could generalize to handle data with inter-span relations, like corpora using eRST or those available at www.aifdb.org

Learning about informal fallacies and the detection of fake news: An experimental intervention The philosophical concept of informal fallacies–arguments that fail to provide sufficient support for a claim–is introduced and connected to the topic of fake news detection. We assumed that the…

Our reading group dissected an experimental study by Hruschka and Appel (doi.org/10.1371/journal.pone.0283238) that underpinned a paper on Persuasion-Augmented Chain of Thought presented at ACL in Vienna (doi.org/10.18653/v1/2025.acl-long.1215).

At this week's reading group, we discussed the paper "Why human–AI relationships need socioaffective alignment" to look at the broader topic of social and emotional dynamics of human-AI interaction, and what this might mean for our own work. www.nature.com/articles/s41...

Practical Solutions to Practical Problems in Developing Argument Mining Systems Debela Gemechu, Ramon Ruiz-Dolz, John Lawrence, Chris Reed. Proceedings of the 12th Argument mining Workshop. 2025.

Debela Gemechu, Ramon Ruiz-Dolz, John Lawrence and Chris Reed: Practical Solutions to Practical Problems in Developing Argument Mining Systems
aclanthology.org/2025.argmini...

The Open Argument Mining Framework Debela Gemechu, Ramon Ruiz-Dolz, Kamila Górska, Somaye Moslemnejad, Eimear Maguire, Dimitra Zografistou, Yohan Jo, John Lawrence, Chris Reed. Proceedings of the 63rd Annual Meeting of the Association…

Debela Gemechu, Ramon Ruiz-Dolz, Kamila Górska, Somaye Moslemnejad, Eimear Maguire, Dimitra Zografistou, Yohan Jo, John Lawrence, Chris Reed: The Open Argument Mining Framework
aclanthology.org/2025.acl-dem...

Mining Complex Patterns of Argumentative Reasoning in Natural Language Dialogue Ramon Ruiz-Dolz, Zlata Kikteva, John Lawrence. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Ramon Ruiz-Dolz, Zlata Kikteva, John Lawrence: Mining Complex Patterns of Argumentative Reasoning in Natural Language Dialogue
aclanthology.org/2025.acl-lon...

Natural Language Reasoning in Large Language Models: Analysis and Evaluation Debela Gemechu, Ramon Ruiz-Dolz, Henrike Beyer, Chris Reed. Findings of the Association for Computational Linguistics: ACL 2025. 2025.

Debela Gemechu, Ramon Ruiz-Dolz, Henrike Beyer and Chris Reed: Natural Language Reasoning in Large Language Models: Analysis and Evaluation
aclanthology.org/2025.finding...

Lexical Recall or Logical Reasoning: Probing the Limits of Reasoning Abilities in Large Language Models Henrike Beyer, Chris Reed. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Henrike Beyer and Chris Reed: Lexical Recall or Logical Reasoning: Probing the Limits of Reasoning Abilities in Large Language Models
aclanthology.org/2025.acl-lon...

CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining Debela Gemechu, Chris Reed. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Find all papers in the ACL Anthology:
Debela Gemechu and Chris Reed: CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining: aclanthology.org/2025.acl-lon...

We're looking forward to a busy and exciting week for our researchers presenting their amazing work at ACL 2025 in Vienna! 🎡
If you're not sure where and when to find us, check the schedule below! 👇

Read the paper here:
📄https://aclanthology.org/2025.acl-long.969/

Read the paper here:
📄https://aclanthology.org/2025.acl-long.664/

Natural Language Reasoning in Large Language Models: Analysis and Evaluation Debela Gemechu, Ramon Ruiz-Dolz, Henrike Beyer, Chris Reed. Findings of the Association for Computational Linguistics: ACL 2025. 2025.

Find the paper here:
📄https://aclanthology.org/2025.findings-acl.192/

Many congratulations to @jooleahd.bsky.social who this afternoon successfully defended her PhD thesis, passing summa cum laude.

Read the paper here:
📄https://aclanthology.org/2025.findings-acl.192/

Read the paper here:
📄https://aclanthology.org/2025.acl-long.368/