SenseDisclosure: A new procedure for dealing with problematically ambiguous terms in cross-disciplinary communication. (September 2018)

• Record Type: Journal Article
• Authors: Mennes, Julie
• Is Part Of: Language sciences. Volume 69(2018)
• Page Start: 57
• Page End: 67
• Publication Date: 2018-09
• Subjects: Cross-disciplinarity -- Interdisciplinarity -- Interdisciplinary communication -- Terminological ambiguity -- Natural language processing -- Word sense induction; Linguistics -- Periodicals; Language and languages -- Periodicals; Linguistique -- Périodiques; Langage et langues -- Périodiques; Language and languages; Linguistics; Periodicals; Electronic journals; 405
• DOI: 10.1016/j.langsci.2018.06.003
• Languages: English
• ISSNs: 0388-0001

Abstract

Abstract: Communication in cross-disciplinary (and thus in inter-, multi- and transdisciplinary ) projects is frequently challenged by problematically ambiguous terms (henceforth 'PATs'), i.e. terms that have multiple meanings and for which it is not always clear what meaning is used, thereby generating communication problems. The reason why communication in cross-disciplinary projects is so sensitive to PATs, is that they often involve disciplines that share one or more terms, yet attribute different meanings to them in an implicit and/or unsystematic manner. The teams of such projects are in need of a PAT resolution procedure, i.e. a procedure that helps them to identify and resolve PATs, as they are generally not trained to do this themselves. A first attempt to provide such a procedure consists in the identification of existing candidate procedures and an evaluation of their capacity to resolve PATs in cross-disciplinary communication contexts using a new set of task and performance criteria. It is shown that none of them sufficiently meets all criteria. It also becomes clear that the realization of an efficient PAT resolution procedure requires the ability to automatically process large quantities of linguistic data. Hence, input from the field of applied (computational) linguistics seems necessary. With this need for automation in mind and against the background of the new set of task and performance criteria, a theoretical characterisation of a new PAT resolution procedure called 'SenseDisclosure' is presented. SenseDisclosure is meant to be applicable to all kinds of cross-disciplinary projects (by an external facilitator). Its characterisation incorporates multiple techniques from Natural Language Processing to realize several critical automations. As the techniques were not specifically developed for PAT resolution, some of them require further research and development before they can be reliably integrated. Finally, it is argued that, if this extra research and development yields positive results, SenseDisclosure can be a truly effective PAT resolution procedure. Highlights: Cross-disciplinary communication suffers from problematically ambiguous terms. No existing procedure resolves problematically ambiguous terms. New computational linguistic procedure resolves problematically ambiguous terms.