Indução de modelos de árvore de decisão para reconhecimento de entidades nomeadas na literatura biomédica

Abstract

The increasing amount of textual documents, especially those related to biomedical literature, has encouraged many researches in Text Mining. One important field of investigation relates to Named Entities Recognition (NER), where Named Entities (NE) are referred terms or objects in a given context. In the biomedical domain, diseases and treatments can be cited as examples of NE. The recognition of biomedical NE has become a challenge, since biomedical corpora have particular characteristics, mainly because a given biological object can be often represented in different terminological ways. Among the different methods of NER, one of them is the recognition through the context. In this work is proposed a Decision-Tree Model-based approach for NER in biomedical literature.