The Possible, the Plausible, and the Desirable: Event-Based Modality Detection for Language Processing

Modality is the linguistic ability to describe events with added information\nsuch as how desirable, plausible, or feasible they are. Modality is important\nfor many NLP downstream tasks such as the detection of hedging, uncertainty,\nspeculation, and more. Previous studies that address modality detection in NLP\noften restrict modal expressions to a closed syntactic class, and the modal\nsense labels are vastly different across different studies, lacking an accepted\nstandard. Furthermore, these senses are often analyzed independently of the\nevents that they modify. This work builds on the theoretical foundations of the\nGeorgetown Gradable Modal Expressions (GME) work by Rubinstein et al. (2013) to\npropose an event-based modality detection task where modal expressions can be\nwords of any syntactic class and sense labels are drawn from a comprehensive\ntaxonomy which harmonizes the modal concepts contributed by the different\nstudies. We present experiments on the GME corpus aiming to detect and classify\nfine-grained modal concepts and associate them with their modified events. We\nshow that detecting and classifying modal expressions is not only feasible, but\nalso improves the detection of modal events in their own right.\n