Creació d'una base de dades de semàntica oracional i un entorn de consulta i gestió (SENSEM)

The present project aims to construct a database (DB) that will collect information about verbs and their behaviour within a sentence. To that purpose, it becomes necessary to build an application tool for the creation, administration and productive consultation of the linguistic database. The unit of description in the DB is the sense and the concept of "verb" will be described from different perspectives: the semantic relations that are established between verbs and their arguments, the syntactic structures in which a verb can be found (sub-categorization frames SCF) the semantics associated to them, verbs selection preferences and, finally, statistics about the frequencies of use of the SCF and the regularities in the co-occurrence of verbs and specific entities. Moreover, examples extracted form a corpus will be linked to all the possible SCF observed in the corpus. The empirical methods upon which the processes of acquisition used are based allow us to account for a relatively large quantity of data for this type of lexicon. All the information will be available for consultation through an interface that provides the user with a view of the syntactico-semantic description of verbs, together with examples of use. Apart from this consultative functionality, the type of resource we propose here will serve as an essential tool for those natural language processing applications that seek to process language at a deep level