An Integrative Approach to Chinese Named Entity Recognition

This paper demonstrates an effective integrative way to recognize the Chinese Named Entities(NEs) in Chinese text, which extends the Maximum Matching and Second-Maximum Matching (MMSM) model while including the POS (Part-of-speech) in the segmentation directed graph. Simultaneously, several Chinese Named Entity Recognition (NER) automata are constructed to recognize all the candidate NEs in Multi-Dimensional instead of single-dimensional context spaces in the directed graph. The Dijkstra algorithm is revised to find the final path in the directed graph with all the candidates NEs and common words. Thus, the recognition of all NEs is accomplished synchronously in an integrative way.