Argument Mining

AMICA ranks COVID-19 papers by query relevance and by richness in claims and evidence, and its ranking agrees with expert opinion.

An argument mining tool scores how much COVID-19 papers argue, and these scores agree fairly to moderately with Cochrane decisions and expert reviewers.

Demosthenes is a corpus of 40 EU Court of Justice decisions on fiscal state aid, annotated with premises, conclusions, and legal argument schemes.

TC-GNN constrains graph pooling to pick out tree-kernel fragments of a sentence parse, and uses them to detect claims and premises across four corpora.

Logic Tensor Networks train argument mining models to fit the data and satisfy domain rules, which improves classification, rule compliance, and robustness on scientific abstracts.

The first system that detects claims in spoken political debates shows that voice features from the audio can improve text-based claim detection.

Argument mining has recently become a hot topic, attracting the interests of several and diverse research communities, ranging from …

A tree kernel over parse trees spots sentences that contain claims without knowing the topic, and performs comparably to systems that rely on topic knowledge.