Robust Argumentation Machines First International Conference, RATIO 2024, Bielefeld, Germany, June 5–7, 2024, Proceedings
This open access book constitutes the proceedings of the First International Conference on Robust Argumentation Machines, RATIO 2024, which took place in Bielefeld, Germany, during June 5-7, 2024. The 20 full papers and 1 short paper included in the proceedings were carefully reviewed and selected f...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | eBook |
| Language: | English |
| Published: |
Cham
Springer Nature
2024
Springer |
| Edition: | 1 |
| Series: | Lecture Notes in Computer Science; Lecture Notes in Artificial Intelligence |
| Subjects: | |
| ISBN: | 3031635361, 9783031635359, 3031635353, 9783031635366 |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- Intro -- Preface -- Organization -- Contents -- Argument Mining -- Natural Language Hypotheses in Scientific Papers and How to Tame Them -- 1 Introduction: Scientific Hypotheses as Complex Claims -- 2 Related Work -- 2.1 Argumentation Modeling for Complex Scientific Claims -- 2.2 Knowledge Representation: Modeling Scientific Language with Knowledge Graphs -- 2.3 Hypothesis Representation in Invasion Biology -- 3 Example: The Biotic Resistance Hypothesis -- 4 Towards Formalizing Scientific Hypotheses -- 4.1 A Generic Structure for Scientific Hypotheses -- 4.2 Linking Hypothesis Formulations to Semantic Models -- 4.3 Classifying Relationships Between General and Specific Claims -- 5 Applications of the Framework -- 6 Limitations -- 7 Conclusions and Outlook -- Appendix -- References -- Weakly Supervised Claim Localization in Scientific Abstracts -- 1 Introduction -- 2 Background -- 2.1 Scientific Claim Detection -- 2.2 Input Optimization for Model Interpretability -- 3 Datasets -- 3.1 The INAS Dataset -- 3.2 The SciFact Dataset -- 4 Method -- 4.1 Span-Level Claim Evidence Localization -- 4.2 Sentence-Level Claim Evidence Localization -- 4.3 Evidence Injection -- 5 Experiments -- 5.1 Span-Level Claim Localization -- 5.2 Sentence-Level Claim Localization -- 6 Results -- 6.1 Span-Level Evidence Localization -- 6.2 Sentence-Level Evidence Localization -- 7 Conclusion -- A Experimental Details -- References -- Argument Mining of Attack and Support Patterns in Dialogical Conversations with Sequential Pattern Mining -- 1 Mining Interactions in Debates -- 2 Related Work -- 3 Predicting a Conversational Dataset -- 3.1 Corpus Creation -- 3.2 Mining Conversation Chains from Incomplete Graphs -- 3.3 Argument Abstraction by Stance and Aspect Prediction -- 3.4 Sequential Pattern Mining on Predicted Data -- 4 Results -- 4.1 Attack and Support Patterns
- 3.1 Overview of the BARD Project and ``the Spider'' Problem -- 3.2 Results with the Original Algorithm -- 3.3 Diagnosis and Solution Proposal -- 3.4 Results of the Improved Version -- 4 Limitation and Future Work -- 5 Conclusion -- A Appendix -- References -- ``Do Not Disturb My Circles!'' Identifying the Type of Counterfactual at Hand (Short Paper) -- 1 Introduction -- 1.1 Introductory Example -- 2 Preliminaries and Related work -- 3 Backtracking in Causal Models -- 3.1 When Backtracking is not Enough -- 3.2 Iterative Backup -- 3.3 Default Logic -- 3.4 Integration of Hyperreals -- 4 Discussion -- References -- Interactive Argumentation, Recommendation and Personalization -- BEA: Building Engaging Argumentation -- 1 Introduction -- 2 Related Work -- 2.1 Argumentative Dialog Systems -- 2.2 Reflective Engagement -- 2.3 Conversational User Engagement and Virtual Avatars -- 3 Prototype and Architecture of BEA -- 3.1 System Architecture -- 3.2 User Interface -- 4 Modeling Reflective Engagement -- 5 Evaluation -- 5.1 Study 1 ch17weber2023fostering: Analyzing Focus on Challenger Arguments -- 5.2 Study 2 ch17aicherspsiva: Influence of Avatar Interface -- 6 Limitations -- 7 Conclusion and Future Work -- References -- Deciphering Personal Argument Styles - A Comprehensive Approach to Analyzing Linguistic Properties of Argument Preferences -- 1 Introduction -- 2 Background -- 2.1 Argument Data -- 2.2 Argument Preferences -- 2.3 Visual Analytics for Linguistics -- 3 The CUEPAQ Argument Exploration Pipeline -- 3.1 The CUEPipe Workflow -- 3.2 Generating a Data Set for Exploring Argument Preferences -- 3.3 Learning Preferences via Visual Interactive Labeling -- 3.4 Exploring Personal Preferences -- 4 Study: Propositional Attitudes -- 5 Limitations -- 5.1 The CUEPipe -- 5.2 The Proof-of-concept Study -- 6 Conclusion -- References -- Argument Search and Retrieval
- 4.4 Batched Prompting -- 5 Experimental Evaluation -- 5.1 Experimental Setup -- 5.2 Datasets -- 5.3 Results and Discussion -- 5.4 Qualitative Error Analysis -- 6 Limitations -- 7 Conclusion and Future Work -- A Prompting Templates -- A.1 Isolated Prompting -- A.2 Sequential Prompting -- A.3 Contextualized Prompting -- A.4 Batched Prompting -- References -- Argument Acquisition, Annotation and Quality Assessment -- Are Large Language Models Reliable Argument Quality Annotators? -- 1 Introduction -- 2 Related Work -- 2.1 Evaluating Argument Quality -- 2.2 LLMs as Annotators -- 3 Experimental Design -- 3.1 Expert Annotation -- 3.2 Novice Annotation -- 3.3 Models -- 3.4 Prompting -- 4 Results -- 4.1 Consistency of Argument Quality Annotations -- 4.2 Agreement Between Humans and LLMs -- 4.3 LLMs as Additional Annotators -- 5 Conclusion -- 6 Limitations -- References -- The Impact of Argument Arrangement on Essay Scoring -- 1 Introduction -- 2 Related Work -- 3 Data -- 3.1 Argument-Annotated Essays Corpus -- 3.2 Feedback Corpus -- 3.3 International Corpus of Learner English -- 4 Experiments -- 4.1 ADU and Sematic Type Classification -- 4.2 Predicting Essay Quality with Flows of Semantic Types -- 4.3 Analysis of Feature Impact -- 5 Discussion -- 6 Conclusion -- References -- Finding Argument Fragments on Social Media with Corpus Queries and LLMs -- 1 Introduction -- 2 Argumentative Fragments -- 2.1 An Inventory of Logical Patterns -- 2.2 Nested Patterns -- 3 Data -- 3.1 Corpus and Linguistic Annotation -- 3.2 Manual Annotation of Argument Fragments -- 4 Corpus Queries -- 4.1 Methods -- 4.2 Evaluation and Discussion -- 5 Hierarchical Queries -- 5.1 Methods -- 5.2 Evaluation -- 5.3 Discussion -- 6 Fine-Tuning LLMs -- 6.1 Methods and Evaluation -- 6.2 Discussion: Qualitative Comparison of Approaches -- 7 Limitations -- 8 Conclusion -- References
- 4.2 Pattern Mining Vs. Analyzing Distributions -- 5 Conclusion -- 5.1 Limitations -- 5.2 Future Work -- References -- Cluster-Specific Rule Mining for Argumentation-Based Classification -- 1 Introduction -- 2 Background -- 3 Cluster-Specific Rule Mining -- 4 Experimental Analysis -- 5 Limitations -- 6 Conclusion -- References -- Debate Analysis and Deliberation -- Automatic Analysis of Political Debates and Manifestos: Successes and Challenges -- 1 Introduction -- 2 Fine-Grained Analysis of Political Discourse -- 2.1 Less Annotation Is More: Few-Shot Claim Classification -- 2.2 Improving Claim Classification with Hierarchical Information -- 2.3 Multilingual Claim Processing -- 2.4 Robust Actor Detection and Mapping -- 3 Coarse-Grained Analysis of Political Discourse -- 3.1 Ideological Characterization -- 3.2 Policy-Domain Characterization -- 4 Conclusions -- References -- PAKT: Perspectivized Argumentation Knowledge Graph and Tool for Deliberation Analysis 5540801En6FigaPrint.eps -- 1 Introduction -- 2 A Data Model for Perspectivized Argumentation -- 3 Constructing PAKTDDO from debate.org -- 3.1 Arguments from debate.org -- 3.2 Characterizing Arguments for Perspectivized Argumentation -- 3.3 Authors and Camps -- 3.4 Implementation and Tools for Building and Using PAKT -- 3.5 Preliminary Evaluation -- 4 Analytics Applied to PAKTDDO -- 5 Case Studies -- 5.1 Should Animal Hunting Be Banned? -- 5.2 Comparison to Other Issues -- 5.3 Argument Level -- 6 Related Work -- 7 Conclusion -- References -- PolArg: Unsupervised Polarity Prediction of Arguments in Real-Time Online Conversations -- 1 Introduction -- 2 Foundations -- 2.1 Computational Argumentation -- 2.2 Natural Language Processing -- 2.3 Online Conversation Platforms -- 3 Related Work -- 4 Prompting Strategies -- 4.1 Isolated Prompting -- 4.2 Sequential Prompting -- 4.3 Contextualized Prompting
- Extending the Comparative Argumentative Machine: Multilingualism and Stance Detection
- Computational Models of Argumentation -- Enhancing Abstract Argumentation Solvers with Machine Learning-Guided Heuristics: A Feasibility Study -- 1 Introduction -- 2 Preliminaries -- 3 Solution Approaches in Abstract Argumentation -- 4 Machine Learning-Guided Heuristics -- 5 Experimental Analysis -- 5.1 Datasets and Setup -- 5.2 Initial Experimental Analysis -- 5.3 Evaluation and Results -- 6 Limitations -- 7 Conclusion -- References -- Ranking Transition-Based Medical Recommendations Using Assumption-Based Argumentation -- 1 Introduction -- 2 Preliminaries -- 2.1 Abstract Argumentation Frameworks -- 2.2 Ranking-Based Semantics -- 2.3 Assumption-Based Argumentation Frameworks -- 3 Ranking Assumptions -- 4 Case Study -- 5 Related Work -- 6 Limitations -- 7 Conclusion -- References -- Argumentation-Based Probabilistic Causal Reasoning -- 1 Introduction -- 2 Preliminaries -- 3 Causal Reasoning -- 3.1 Defeasible Causal Reasoning -- 3.2 Probabilistic Causal Reasoning -- 4 Counterfactual Reasoning -- 5 Discussion -- 6 Limitations -- 7 Conclusion -- References -- From Networks to Narratives: Bayes Nets and the Problems of Argumentation -- 1 Introduction -- 2 The Bayesian Approach to Argumentation -- 2.1 The Bayesian Framework -- 2.2 Bayesian Belief Networks (BBNs) -- 2.3 Explaining BBNs: Important Challenges -- 3 Algorithmic Approaches to Bayesian Argumentation -- 3.1 The Relation Between Argument Diagrams and Bayesian Networks -- 3.2 Introducing Three Extant Algorithms -- 3.3 Evaluating the Algorithms: Example Networks -- 4 Limitation -- 5 Conclusion -- References -- Enhancing Argument Generation Using Bayesian Networks -- 1 Introduction -- 2 The Question of Independent Arguments -- 2.1 Factor Graphs -- 2.2 Overview of the Factor-Graph-Approach Proposed by J. Sevilla -- 3 Testing and Improving the Factor Graph Algorithm