New arXiv Paper Presents Joint Argument-Entity Detection in Political Debates

A new preprint introduces a joint approach to detecting and classifying both arguments and the entities they involve in political debate texts. The authors argue that claims and premises are hard to interpret on their own, so linking them to relevant entities makes argument mining outputs more meaningful. The work appears on arXiv's Computation and Language category.