An Italian question answering system for structured data based on controlled natural languages

Question Answering over structured data represents one of the main challenges in the field of Natural Language Processing since it requires to render natural language, which is used by people every day, into a formal language, which can be processed by a machine. This task is particularly tricky due to the gap between the vocabularies adopted by users and the formalism that characterizes any query language. For this reason, although its birth as a discipline dates back to the late sixties, Question Answering over structured data is still accomplished to an unsatisfying degree. This result is even more critical if we take into account languages different from English, for which the amount of available resources is limited. In this paper we present MULIB, a Question Answering system capable of answering questions in Italian over both Knowledge Bases and databases.