Abstract

Chinese proper nouns recognition (CPNR) plays an important role in the fields of Information Extraction, Question Answering and Text Mining. In this paper we proposed a novel pattern-matching-based method to recognize proper nouns (PNs), which includes person names, location names and organization names, and mainly conducted the following studies: (1) constructing the PNs inner-pattern set; (2) acquiring the PNs outer-patterns by clustering and evaluating automatically; (3) resolving the conflicts of the PNs recognition by computing PNs reliability; (4) conducting the experiment on 1.2M word corpus that are chosen from People Daily corpora. The experimental results, whose recall and precision are 83.4% and 80.1% respectively, indicate that the proposed method is feasible and effective.

Paper Details