Abstract

The semantic analysis of verbs is a theoretical and methodological challenge due to the syntactic and semantic complexity of these lexical units. The aim of this research is to identify the lexical-syntactic structures, that is, the patterns formed considering the syntactic structure, the arguments and the semantic types of the Spanish verbs. The analysis of Spanish verbs is carried out following the proposal of Corpus Pattern Analysis (Hanks, 2004a). This analysis is complemented with the automation of the procedure, combining a dependency analyser with a series of algorithms based on corpus statistics. As a result of the process, a database of lexical-syntactic patterns of 182 manually annotated verbs is offered, as well as an interface for automatic analysis which, according to the evaluation carried out, shows a 63.41% accuracy with respect to manual identification. This would be a contribution both to the theory of lexical semantics and to the description of the Spanish lexicon from a corpus-based methodology.