Support Vector Machine based classification system for classification of sport articles

Support Vector Machine (SVM) is a classification technique used for the classification of linear as well as non-linear data. SVM is the margin based classifier. It selects the maximum margin. In this paper, we present SVM based classification system that classify the given sport articles as cricket relevant and other sport using SVM Light tool. Sport articles in the form of text documents are first converted into a format suitable for SVM Light. Based on training data, SVM Light builds the SVM model. This model is further used to perform classification of testing data. On the basis of result of classification, the confusion matrix for the classifier is discussed, The total number documents related to cricket and other sport from test data is also displayed.