Domain Classification of Textual Conversation Using Machine Learning Approach

This paper presents an approach for classification of textual conversation into multiple domain categories using support vector classifier. The feature reduction is done through Principal Component Analysis (PCA) to extract the important features from the feature vector. These features are passed to different configurations of SVM and the best one is chosen for the final process of classification. The domain's categories are defined on real life situations and conversation to train the system like education & research, personal, patriotism, terrorism, medical, religious, sports, and business. The experiment results show that the proposed method works effectively with more than 75% accuracy.