The PAROLE model and the Italian Syntactic lexicon

Abstract - This paper presents an overview of a large scale Syntactic Computational Lexicon of Italian. This lexicon was elaborated in the framework of the EC funded LE-PAROLE project, which developed core, generic and re-usable written language resources in 12 EU languages. All monolingual lexica were built according to the same design principles, same linguistic specifications and representation format. The PAROLE Italian lexicon is representative of modern Italian language use. The entries were selected on a frequency basis from the ILC Corpus and the syntactic structures encoded were partly inferred from their contexts of occurrence. Both the general structure of a PAROLE lexicon and the specificity of its Italian instantiation are presented. Some languagespecific linguistic and lexicographic options concerning crucial issues to a lexicon building process are illustrated. An overview of the syntactic structures encoded for verbs, nouns and adjectives allows lexicon syntactic coverage as well as description fine-grainedness to be estimated.