CRF+LG: A Hybrid Approach for the Portuguese Named Entity Recognition

Abstract

Named Entity Recognition is an important and challenging task of Information Extraction. Conditional Random Fields (CRF) is a probabilistic method for structured prediction, which can be used in the Named Entity Recognition task. This paper presents the use of Conditional Random Fields for Named Entity Recognition in Portuguese texts considering an additional feature informed by a Local Grammar. Local grammars are handmade rules to identify named entities within the text. Moreover, we also present a study about the boundaries of CRF’s performance when using a result coming from any other classifier as an additional feature. Two well-known collections in Portuguese were used as training and test sets respectively. The results obtained outperform results of state-of-the-art systems reported in the literature for the Portuguese.