Chinese Named Entity Recognition via Joint Identification and Categorization

Chinese Named entity recognition (NER) is an important task for Chinese information processing. Traditional sequence labeling approaches to Chinese NER cannot treat globally a string of continuous characters as a named entity candidate so that the entity-level features cannot be exploited in a natural way. To deal with this problem, we formulate Chinese NER as a joint identification and categorization task that performs the two subtasks simultaneously: boundary identification and entity categorization, together with segmentation. The proposed approach provides a natural formulation to treats pieces of continuous characters as named entity candidates, which allows for more accurate prediction by examining both the internal evidence and contextual information of the candidates. Within this framework, we explored a variety of effective feature representations for Chinese NER. Closed tests on two quite different corpora from the third SIGHAN bakeoff show that our approach significantly outperforms the best in the literature, achieving state-of-the-art performance.