Based cascaded conditional random fields model for Chinese Named Entity recognition

This paper presents a new approach of Chinese named entity recognition based on cascaded conditional random fields. In the proposed approach, The model structure has been designed with the cascade way, the result then is passed to the high model and suppose the decision of high model for recognition of the complicated organization names. Person and location were recognized using firstly rule-based and lastly statistical-based, which is different from the previous BIO label recognition approach. But, the organization recognition is recognized using firstly statistical-based and lastly rule-based. Some interesting features have been proposed, the new probabilistic feature is proposed, which are used instead of binary feature functions, however, it is one of the several differences between this model and the most of the previous CRFs-based model. We also explore several new features in our model, which includes confidence functions, position of features etc. We evaluate our approach on large-scale corpus with open test method using Peoplepsilas Daily (January, 1998), The evaluation results show that our approach based on cascaded conditional random fields significantly outperforms previous approaches.