A text generation module for a decision support system (natural language, computational linguistics)
Ping-Yang Li · 1985
Our text generator is one module of a decision support system designed to assist physicians in the management of stroke, which is being developed jointly by the Neurology Department at Michael Reese Hospital and the Computer Science Department at Illinois Institute of Technology, using data from the Michael Reese Stroke Data Base. The text generator produces multi-paragraph reports on stroke cases stored in the Stroke Data Base or on cases being processed by the Decision Support System. Analysis of human-generated case reports using Sager's Linguistic String Parser led to a characterization of the stroke sublanguage in terms of four components: a Text Grammar for stroke case reports, a set of Stroke Information Formats, a Relational Lexicon for the stroke sublanguage, and a Linguistic String Grammar for this sublanguage. Our first reports were highly constrained; the program chooses between alternative formulas depending on the symptoms input by the physician and deductions from that information. We are now producing freer text by using reverse transformations from our LSP grammar to combine fragments into sentences. Further interest lies in discovering how to generate good paragraphs, using the Text Grammar, the Stroke Information Formats, the Relational Lexicon, and the Linguistic String Grammar as tools.