QADiscourse -- Discourse Relations as QA Pairs: Representation, Crowdsourcing and Baselines
This addresses the challenge of requiring expert annotators for discourse relation identification in natural language understanding, offering a more accessible and scalable approach.
The paper tackles the problem of annotating discourse relations by proposing a novel representation as QA pairs, enabling crowdsourcing of a wide-coverage dataset and achieving baseline prediction results.
Discourse relations describe how two propositions relate to one another, and identifying them automatically is an integral part of natural language understanding. However, annotating discourse relations typically requires expert annotators. Recently, different semantic aspects of a sentence have been represented and crowd-sourced via question-and-answer (QA) pairs. This paper proposes a novel representation of discourse relations as QA pairs, which in turn allows us to crowd-source wide-coverage data annotated with discourse relations, via an intuitively appealing interface for composing such questions and answers. Based on our proposed representation, we collect a novel and wide-coverage QADiscourse dataset, and present baseline algorithms for predicting QADiscourse relations.