یک طرح برنامه نویسی شبکه ی Bayesian برای حاشیه نویسی اطلاعات زیست پزشکی ارائه شده به مشتریان مشاوره ژنتیک

We developed a Bayesian network coding scheme for annotating biomedical content in layperson-oriented clinical genetics documents. The coding scheme supports the representation of probabilistic and causal relationships among concepts in this domain, at a high enough level of abstraction to capture commonalities among genetic processes and their relationship to health. We are using the coding scheme to annotate a corpus of genetic counseling patient letters as part of the requirements analysis and knowledge acquisition phase of a natural language generation project. This paper describes the coding scheme and presents an evaluation of intercoder reliability for its tag set. In addition to giving examples of use of the coding scheme for analysis of discourse and linguistic features in this genre, we suggest other uses for it in analysis of layperson-oriented text and dialogue in medical communication.