Analysing clinical guidelines' contents with deontic and rhetorical structures

G. (Gersende) Georg, H. (Hugo) Hernault, M. O. (Marc) Cavazza, H. (Helmut) Prendinger, M. (Mitsuru) Ishizuka

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Original languageEnglish
Title of host publicationLecture notes in computer science
ISBN (Print)9783642029752
Publication statusPublished - 2009

Bibliographical note

The computerisation of clinical guidelines can greatly benefit from the automatic analysis of their content using Natural Language Processing techniques. Because of the central role played by specific deontic structures, known as recommendations, it is possible to tune the processing step towards the recognition of such expressions, which can be used to structure key sections of the document. In this paper, we extend previous work on the automatic identification of guidelines' recommendations, by showing how Rhetorical Structure Theory can be used to characterise the actual contents of elementary recommendations. The emphasis on causality and time in RST proves a powerful complement to the recognition of deontic structures and supports the identification of relevant knowledge, in particular for the identification of conditional structures, which play an important role for the subsequent analysis of recommendations.

Cite this