基于支持向量机的中文分词  

Chinese Word Segmentation Based on Support Vector Machine

作　　者：林秋虾[1] 

机构地区：[1]华侨大学厦门工学院计算机科学与工程系,厦门361021

出　　处：《现代计算机》2011年第19期11-13,共3页Modern Computer

摘　　要：中文分词是中文信息处理的基础,也是很多中文应用首先要面对的问题。目前效果最好的分词模型是词位标注法,该方法需要使用一个分类器对每个汉字的位置进行判定。基于统计学习理论的支持向量机较好地解决小样本、非线性、高维数和局部极小点等实际问题,被公认为是优秀的机器学习方法和分类算法。实现一个基于支持向量机的中文分词系统,并在实验中取得较好的结果,证明支持向量机适用于中文分词时的词位标注问题。Chinese word segmentation is the basis for Chinese information processing, and is often the first problem a lot of Chinese applications must solve. Currently the best segmentation model is the word-position tagging method, which requires a classifier to determine the location of each character. Support vector machine based on statistical learning theory has already proved it can solve the small sample, nonlinear, high dimension and local minimum point of practical problems.So it is recognized as one of the best machine learning methods and classification algorithms. Realizes a Chinese word segmentation system based on support vector machine. The experiment achieves a good result. So it proves that support vector machine is effective for word-position tagging problem.

关 键 词：中文分词 词位标注 支持向量机 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]