关键词: 支持向量机 基于转换的错误驱动学习 汉语组块识别

Abstract:

The paper presents a method of Chinese chunk recognition based on Support Vector Machines（SVM） and transformation-based error-driven learning. It is well known that SVM is good at selecting useful features and achieving high generalization of very high dimensional feature space. Furthermore, by introducing the kernel function, SVM can be trained in a high dimensional space with smaller computational cost independent of their dimensionalities. The t ransformation-based learning approach with supervision is further applied to improving the analysis results of SVM. Transformation rules effectively dea l with the specificity of Chinese and improve the performance of SVM. The experiments show encouraging results.

Key words: support vector machines（SVM）, transformation-based error-driven learning;Chinese chunk recognition