ME-BASED CHINESE PERSON NAME AND LOCATION NAME RECOGNITION MODEL

摘要：

This paper constructs a hybrid model for automatic Chinese person name and location name recognition, which is based on Maximum Entropy principle.The model consists of a training module and a recognizing module.Firstly, contextual features are extracted from the training corpus.Maximum Entropy principle is employed to train the features.Then, the trained features together with a Dynamic Word List and a simple Rule Base are used to recognize Chinese person names and location names in the testing corpus.The experimental results are satisfying and have been analyzed.

关键词： Maximum entropy model Named entity recognition Feature extraction Linguistic rules

作者: YUE-JIE ZHANG TAO ZHANG

作者单位: Department of Computer Science and Engineering, Shanghai Key Laboratory of Intelligent Information P School of Information Management and Engineering, Shanghai University of Finance and Economics, Shan

会议类型: 国际会议

会议名称: 2007 International Conference on Machine Learning and Cybernetics(IEEE第六届机器学习与控制论国际会议)

会议地点: 香港

会议语种:英文

页码: 3442-3447

在线出版日期: 2007-08-19（万方平台首次上网日期，不代表论文的发表时间）