Evaluation of Supervised Classification Techniques on Twitter Data using R

The phenomenal development of the World Wide Web has resulted in enormous social networking sites producing tremendous data on web 2.0. Social networking sites have widened to a higher degree of use, in which any field of information can be sort by researchers. Data obtained from social media has strategized from many new machine learning algorithms and natural language processing. The data is unstructured; mining the data leads to finding important sentiments about various entities via appropriate classification techniques. In this paper, tweets’ opinions are analyzed through machine learning algorithms such as naive Bayes and support vector machines using R programming; results are computed and compared. The SVM model manifests the higher precision, and naïve Bayes provides higher accuracy for sentiment analysis on the Bengaluru traffic data.