摘要
文本分类对于各个领域挖掘文本信息非常重要,论文在基于频率核函数的文本分类基础上,充分比较各种分类器的优缺点,提出一种利用稀疏表示分类器(SRC)和支持向量机(SVM)的组合方法进行文本分类,以提高文本分类的性能。最后通过实验表明,使用二者结合的方法效果明显好了很多。
Text classification is very important for various fields of management of a large number of text content,based ontext classification based on the frequency of kernel function,the advantages and disadvantages of all kinds of classifiers are com-pared,this paper proposes a sparse classifier(SRC)and support vector machine(SVM)combination method to improve the perfor-mance of text classification. Sparse representation classifier the field dictionary is constructed by means of the vector representationof the document. SVM text classifier linear kernel function and Gauss kernel function are used for the vector representation of thedocument.
出处
《计算机与数字工程》
2017年第12期2479-2481,2497,共4页
Computer & Digital Engineering
关键词
稀疏表示
SVM
频率核函数
文本分类
sparse representation
SVM
frequency-based kernels
text classification