Kartu-Verbs: A Semantic Web Base of Inflected Georgian Verb Forms to Bypass Georgian Verb Lemmatization Issues

The Georgian language has a complex verbal system, both agglutinative and inflectional, with many irregularities. Inflected forms of a given verb can differ greatly from one another and it is still a controversial issue to determine which lemmas should represent a verb in dictionaries. Verb tables help people to track lemmas starting from inflected forms but these tables are tedious and error-prone to browse. We propose Kartu-Verbs, a Semantic Web base of inflected Georgian verb forms. For a given verb, all its inflected forms are present. Knowledge can easily be traversed in all directions: from Georgian to French and English; from an inflected form to a masdar (a verbal noun, the form that comes closest to an infinitive), and conversely from a masdar to any inflected form; from component(s) to forms and from a form to its components. Users can easily retrieve the lemmas that are relevant to access their preferred dictionaries. Kartu-Verbs can be seen as a front-end to any Georgian dictionary, thus bypassing the lemmatization issues.