一种可训练的快速汉语部分句法分析方法

A Trainable and Fast Partial Parsing Method for Chinese

提出一种可训练的快速汉语部分句法分析方法 .此方法基于文中提出的扩充多层有限状态自动机模型 ,其中引入错误驱动的机器学习方法使其正确率可以通过训练得到提高 ,初步结果表明 ,该方法是行之有效的 .

In this paper, a trainable and fast partial parsing method for Chinese is presented. This method is based on the augmented cascaded finite state automaton proposed in this paper, which uses error driven machine learning to achieve high accuracy. The preliminary result shows that the proposed method is promising.