SIGDOC Graduate Competition
Ashley Kelly, Allan McDougall, and Nike Abbott conducted the work that they are presenting with the Inkpot Natural Language Processing group in the David R. Cheriton School of Computer Science. They were asked to bring their knowledge of rhetorical theory to natural language processing problems; the results of this are what they will be presenting at the 27th ACM International Conference on Design of Communication, in Bloomington, Indiana in October. Their paper will be published in the conference proceedings. Their paper, entitled "Rhetorical Models for Computational Systems: an interdisciplinary approach to reusable, tailorable medical information" won the best graduate student submission award.
The following is the abstract for their paper:
This paper aims to bring rhetorical scholarship to the field of Natural Language Processing (NLP) and further its inclusion in health communication studies. Much work in NLP relies upon statical machine learning techniques, which, while valuable, are still limited in their abilities. Rhetorical theory, the study of suasions, offers a new paradigm for NLP research and development. In this pilot study, they use rhetorical theory as an analytic approach to modeling discourse patterns in a highly tailored corpus of texts. Along with their findings, they argue that rhetorical theory and analysis can further the creation of both tailored documentation and computationally-generated texts.
Special thanks to Professor Neil Randall for his helpful advice and guidance throughout the process.