Real-Time Classification of Twitter Data Using Decision Tree Technique

Abstract

The data which comes from e-commerce site is unstructured text data. Text mining is becoming an important field in research for finding valuable information from unstructured texts. Data which contains an unstructured text which stores large volume of valuable information cannot exclusively be used for any process by computers. Therefore, we want a definite process strategies, techniques and algorithms in order to extract this meaningful information which is done by using text mining. Classification of these user opinions is an Information Extraction and Natural Language Processing task that classifies the user opinions into various categories like in the form of positive or negative. In these paper, we can build a classifiers based on SVM and decision tree classification algorithm to identify the opinions and classify them onto categories, and also, we can compute the performance measure of these classifiers and also compare the classification algorithms performance based on their accuracy, and we can say that decision tree perform better as compared with SVM. In this, we can also classify the real-time tweets review into various emotions and polarity through decision tree classification model.