Design ed implementazione di una metodologia per l'allineamento di risorse semantiche e la popolazione automatica di spazi concettuali.

In this thesis I introduce the TTCS system, that exploits a resource-driven approach relying on BabelNet, NASARI and ConceptNet. TTCS takes in input a term and its context of usage and produces as output a novel type of vector-based semantic representation, where conceptual information is encoded through the conceptual spaces (a framework for common-sense knowledge representation and reasoning). To these ends, each term in input is automatically anchored to the corresponding concept in the underlying representation by adopting as common ground the BabelNet synset ids, thus providing a a uniform language/concept interface. Then, the TTCS provides a common-sense description of the concept at hand by mixing the information retrieved in BabelNet and in ConceptNet. The system has been evaluated through a twofold experimentation aimed at assessing i) the quality of the extracted common-sense conceptual information w.r.t. human judgments; ii) the usefulness of the obtained representations in a wider context. To serve this purpose the TTCS results have been employed in DUAL-PECCS, a previously existent categorization system based on both ontologies and conceptual spaces. In both cases the results are encouraging and provide precious insights to make substantial improvements.