The Latin WordNet 2.0 is an on-line, open-access lexico-semantic database of the Latin language. Building on and extending the original work of Stefano Minozzi, this project, under the direction of William Michael Short at the University of Exeter, aims to create a comprehensive model of Latin's semantic system—including its figurative structures and diachronic development—that can be handled and manipulated computationally. Besides constituting an accurate model of the meanings of Latin words, the WordNet also seeks to include rich information about lexical, semantic, and domain relations, thus representing the language as a densely interconnected conceptual map. The original WordNet data is being curated in collaboration with the LiLa: Linking Latin Project, whilst a further about 70,000 lemmas are being added. A second joint project with the University of Genoa is building a dictionary of conceptual metaphor (Lexicon Translaticium Latinum) as an extension of the Wordnet, in order to capture large-scale figurative patterns in Latin.
Available through an intuitive RESTful API, the WordNet can be viewed on the Web or accessed programmatically for integration into other natural language understanding applications.