ABSTRACT
Support Vector Machines (SVM) has an important position in Machine learning
theory, whether it is to solve the classification problem or request for the reunification
issue, SVM has a wide range of applications. In this paper, a short introduction into the
basic principles of SVM, a detailed discussion of the SVM in the text classification, and
a careful analysis of how to make use of SVM to construct classifier for a text
classification. Here's the text of the detailed classification process and introduced in the
course of these key technologies, such as: segmentation technology, vector space model
(VSM), features selection technology, cross-verification technology of the SVM and so
on. With the analysis and discussion also briefly described the process of making use of
Microsoft Visual C++ 6.0 to create the text classification system, introduced the
realization and optimization of the key class and important functions, and how to use of
dynamic link library to achieve the migration from C++ to Java. Finally, the
experimental data and conclusions produced by this system are shown.
Keywords: machine learning text classification SVM(support vector machine)
