Latin WordNet, una rete di conoscenza semantica per il latino e alcune ipotesi di utilizzo nel campo dell’information retrieval
This paper describes the creation of Latin WordNet, addressing the methods of construction, the data structure, possible applications and further developments. More notably, a possible expansion of Latin WordNet is the construction of an additional data structure in order to improve the chances of the network to represent the semantic shift taking diachrony into account. The original Princeton WordNet, on whose ground Latin WordNet is developed, does not take into account the semantic shift of words in time, considering the meanings identified by synsets as if existing in a model of vocabulary where a language has not an history: the purpose of Princeton WordNet was obviously the representation of in-use contemporary English, but this model is weak when we come to describe the semantic structures of a finished language such as Latin. In the second part, this paper shows the connections between semantic networks and information retrieval moving from the problems pertaining to the automatic disambiguation of contexts and the exploitation of ontologies in the process of sense retrieval on semantically annotated corpora.