"Negation Detection in Automated Medical Applications";
Report for Asgaard-TR-2006-1;
In medical reports patient data is mostly stored in narrative language, which is the spoken or written input from the responsible physicians. To allow computer processing it is necessary to translate this natural language into a format a computer is able to understand and deal with. In this wide area of translation efforts the so called "Negation Detection", i.e. the detection of negated phrases, which do not deliver supplementary information, has an essential importance, since for automatic systems the negation detection can work like a filter system, filtering irrelevant information from important knowledge pieces. It is necessary to decide whether a given medical phenomenon must be taken into account or if it can be ignored, because it is not present at the patient of interest.
This survey examines already existing negation detection procedures and compares their accuracy. Furthermore relevant background information is provided in order to get familiar with the vocabulary used in this special field.
Electronic version of the publication:
Created from the Publication Database of the Vienna University of Technology.