The Research of Semantic Kernel in SVM for Chinese Text Classification
Mai Fan-jin, Ling Huang, Tan Jing, Wang Xinzheng · 2017
The study of semantic kernel function is an important branch of kernel function research in recent years. However, there are few studies on the use of semantic kernel function in support vector machine for Chinese text classification. This paper constructs a domain-related weight matrix based on statistical features and a semantic kernel function based on the combination of "HowNet" and "Synonyms", and how to make full use of semantic relations in Chinese text to improve classification performance. The semantic kernel function is embedded in the support vector machine classifier for Chinese text classification experiment. The test results show that the accuracy, recall rate and F1 value of the support vector machine (SVM) of the semantic kernel function are higher than that of a single ontology or statistic-based semantic kernel function.