IAT2BAS Dataset

The dataset comprises a collection of publicly available Inference Anchoring Theory (IAT)–annotated corpora that have been restructured to derive Bipolar Argument Structures (BAS). At present, the repository includes three diverse corpora adapted into BAS; however, the framework is extensible and can be applied to incorporate all publicly available IAT corpora in the Argument Interchange Format (AIF). These corpora support benchmarking and computational modeling for argument structure prediction across a wide range of multi-party dialogical interactions, grounded in the Inference Anchoring Theory framework. The three corpora shared in this dataset are: QT30 Corpus (Hautli-Janisz A, Kikteva Z, Siskou W, Gorska K, Becker R, Reed C. QT30: A Corpus of Argument and Conflict in Broadcast Debate. In: Proceedings of the Thirteenth Language Resources and Evaluation Conference. Marseille, France: European Language Resources Association; 2022. p. 3291-300.) US2016reddit Corpus (Visser J, Konat B, Duthie R, Koszowy M, Budzynska K, Reed C. Argumentation in the 2016 US Presidential Elections: Annotated Corpora of Television Debates and Social Media Reaction. Language Resources and Evaluation. 2020 Mar;54(1):123-54) RIP1 Corpus (Schad E, Visser J, Reed C. The RIP Corpus of Collaborative Hypothesis-Making. In: Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). Torino, Italia: ELRA and ICCL; 2024. p. 16047-57) All datasets have been accessed via the AIFdb repository. All datasets are credited to their original owners. Each conversation is stored in the following JSON format: {"conversation_id": "string", % original unique identifier of the conversation "conversation_text": "string", % raw conversation text"argument_objects": {"argument_units": [{"id": "integer", %unique identifier of conversation "text": "string", %verbatim argument unit span taken from conversation"reason": "string" %brief explanation of argument unit as per IAT annotations}, ...],"relations": [{"source_id": "integer", %unit ID that is supporting/attacking"target_id": "integer", %unit ID that is being supported/attacked"type": "string" %relation type: support / attack}, ... ]}, ... } For more detailed documentation on our data processing pipeline, please refer to our Github Repository: IAT-BAS-Data-Pipeline