Analysis of bagged ensemble classifiers for blogger data
Muthukumarasamy Govindarajan · 2016
Blogs, review sites, Twitter, and other social networks are the most common platforms that are used by people and organizations for posting their views. People's opinions, emotions, and attitude towards particular aspect can studied through opinion mining. It has been observed through this study that higher accuracy has been achieved by constructing an ensemble using bagging with single classifiers such as Support Vector Machine (SVM), Naive Bayes (NB), Genetic Algorithm (GA) and a comparison evaluation is performed on the blogger dataset which is commonly used for sentiment mining. The experimental results exhibits the improved performance of proposed classifiers over the single classifiers.