UN MODELO DE RECUPERACIÓN DE INFORMACIÓN BASADO EN SVMs INFORMATION RETRIEVAL MODEL BASED ON SVMS

Abstract

Classifiers like the SVMs (Support Vector Machines) were used to organize documents very efficiently, but their effectiveness has not been verified in Information Retrieval for document hierarchizing. This paper proposes a transformation that associates the IR process to a new vectorial space where a SMVs-based classifier is trained to learn the concept of similarity with documents.