ordklasseenhet

ordklasseenhet is a term used in linguistics and corpus linguistics to describe the basic unit that carries a word class (ordklasse) in analysis. It typically refers to a token, such as a word in a text, that is assigned a grammatical category like noun, verb, adjective, or other parts of speech. In many frameworks, an ordklasseenhet also carries morphological information (such as number, tense, or degree) and a lemma, linking the inflected form to its base form.

In practice, the ordklasseenhet is the unit used for part-of-speech tagging and syntactic parsing. An annotated corpus will label each ordklasseenhet with its probable class in context. This labeling supports downstream tasks such as syntactic parsing, machine translation, and information extraction, where understanding a word’s function in a sentence is essential. The exact set of classes can vary by language and annotation scheme, but Norwegian schemes typically include substantiv (noun), verb, adjektiv (adjective), adverb, pronomen (pronoun), preposisjon (preposition), konjunksjon (conjunction), determiner, numeral, and interjeksjon (interjection).

A key characteristic of ordklasseenhet is its context-dependence. Some words can belong to more than one ordklasse depending on usage; for example, a form may function as a verb in one sentence and as a noun in another. Multiword expressions can complicate the concept, as the relevant ordklasseenhet may span more than a single token. Thus, reliable tagging often requires contextual analysis and, in some cases, disambiguation rules or machine learning models.

See also: ordklasse, parts of speech, POS tagging, corpus linguistics, morphosyntactic annotation.