Phrasing analysis is an important item in natural language processing. Identifying Chinese BaseNP can reduce the difficulty of the phrasing analysis processing.The system sort the rules based on the training corpus with POS. and boundary of each BaseNP. By means of extracting the syntactic composition templates, and combining with consulting the contextual information, the system can acquire rules automatically. It provides a self-learning mechanism monitored by manual to process raw corpus, complementing and completing the basic rules. In the process of dentifying, we use the method of maximum matching the rules and give a special solution of some templates. Experiments show that this method is useful and results in higher precision and recall precision. |