Fig. 6From: Negation detection in Dutch clinical texts: an evaluation of rule-based and machine learning methodsIntra-word negation variance. The token delimiter character Ä is a result of the tokenizationBack to article page