Lexical ambiguity detection in professional discourse

Professional discourse is the language used by specialists, such as lawyers, doctors and academics, to communicate knowledge assumptions associated with their respective fields. can be especially difficult for non-specialists understand due lexical ambiguity of commonplace words that have a different or more specific meaning within specialist domain. This phenomena also makes it harder specialists general public because they are similarly unaware potential misunderstandings. In this article, we present an approach detecting domain terms versus everyday English. We demonstrate efficacy our three case studies in statistics, law biomedicine. all studies, identify [email protected] greater than 0.9, outperforming best performing baseline 18.1–91.7%. Most importantly, show ranking broadly consistent semantic differences. Our results highlight difficulties existing difference methods cross-domain setting, which rank non-domain highly noise biases data. additionally generalizes short phrases investigate its data efficiency varying number labeled examples.