A Novel Hybrid Algorithm For Text Classification To Implement Sentiment Analysis

Abstract

The word used by a person has strongly connect with the emotions the words have been analysed and classified as positive, negative and neutral by sentiment analysis if the most positive words used by a person seems to be normal in life if it is negative thinks to be abnormal. The text classification done by the popularly known algorithms like Support vector machine(SVM), Naive Bayes(NB), K Nearest Neighbour(KNN), Decision Tree – J48(DT - J48). By these classification algorithm the text data been analysed and get result accuracy by fitness function based on result accuracy the SVM gives high accuracy and NB in next position. These two algorithms combined together and produce a hybrid algorithm to get more accuracy in text classification.