Towards a Standardized Linguistic Annotation of the Textual Content of Labels in Knowledge Representation Systems