Annotating Mood, Tense and Voice in Czech Corpora

Abstract In the corpora of the Czech National Corpus, the verbtag attribute was introduced for annotating the mood, voice, person, and tense of both simple and compound verb forms. This article presents this attribute and the process of its automatic annotation. We then show an example of the use of verbtag in corpus research by comparing five different text genres in terms of verbal categories expressed in this attribute.