In this article, we study the techniques used for detecting errors in Natural Language Processing (NLP). We classify the techniques according to their approach (symbolic or empirical), and then we describe them in depth. Following that, we describe the systems we have developed for detecting syntactic errors in Basque, by using that technique as a criterion for the classification of those systems, and enhancing it with examples.
© 2008-2024 Fundación Dialnet · Todos los derechos reservados