Usage features. Any attempt at finding a single data vocabulary to capture usage features of senses seems unrealistic. As Chapter 1 has shown, most usage features are indicated through labelling, but the meaning of each label (and how it relates to other labels) is specific to the thesaurus it is found in.69 A label should therefore always be seen within the defined context of its body. As a result, a shared terminology and definitions for these labels is not likely to be found. LexInfo, for instance, defines a formal register but not an informal one. Moreover, the exact relation between its temporal qualifiers – archaic, obsolete, and outdated – is left unspecified and such relations may very well differ between thesauri. Are 67The repository contains four definitions for verb, of which two have the status “standardized”. Further efforts beyond the Semantic Web towards standardizing parts of speech exist, too. A case in point is the Universal Dependencies framework, which employs the CoNLL- U file format to capture information on sentence tokens, including a universal part of speech tag (e.g., verb, noun) and a language-specific one (see ‘CoNLL-U Format’, Universal Dependencies). 68Diller, Review of HTE1, p. 322. 69Atkins and Xxxxxxx, The Oxford Guide to Practical Lexicography, pp. 182–6. outdated items per definition also considered obsolete? Quite as is the case for the part of speech attribute, which vocabulary is best suited to express usage features may depend on the thesaurus in question. If no adequate vocabulary exists, some usage features may even best be represented with terminology specifically coined per thesaurus. One usage feature that sometimes went beyond mere labelling in the existing historical language thesauri was diachronic marking: stating when a particular sense was in use. This temporal aspect is conveyed in thesauri through named, or even dated, periods in time. One particular data vocabulary, found in LOV, appears highly suitable for capturing such temporal aspects: the Time Ontology in OWL (OWL-Time).
