Comparative analysis of ensemble classifiers for sentiment analysis and opinion mining


Ensemble classifiers are showing a promising way to solve various classification and predictive problems. Improving the performance of ensemble classifiers becomes easier than improving the performance of a single classifier. For performance improvement, the comparative analysis of variously available ensemble classifiers is required for better understanding of the ensembles on various kinds of datasets. In this paper one single classifier and three ensemble classifiers are applied on a movie review data set. The results are compared, and it is found that ensemble classifiers are performing way better than single individual classifier. Read More

Publication: 2017 3rd International Conference on Advances in Computing, Communication & Automation (ICACCA) (Fall)

Publisher: IEEE Xplore

Authors: Ravendra Singh & Sanjeev Kumar

Keywords: Ensemble Classifier, Sentiment Analysis, AdaboostMl, Bagging, Stacking

Meet one of the Author:

Dr Sanjeev Kumar received his Ph.D. in Computer Science and Information Technology from M.J.P. Rohilkhand University, Bareilly, Uttar Pradesh, India. His research background encompasses Natural Language Processing, Machine Learning, Deep Learning, Ensemble Classifier, Optimization Algorithms, and Stream Mining. His ability to think critically and approach problems in a logical and analytical manner differentiates him from others.

Dr. Sanjeev Kumar


Department of Computer Science and Information Technology, M.J.P. Rohilkhand University, Bareilly, Uttar Pradesh, India – Mahatma Jyotiba Phule Rohilkhand University (MJPRU) is a university located in the state of Uttar Pradesh, India. It has both affiliated and campus jurisdictions and is governed by the UP Universities Act, 1973. The university offers higher education and has a long history, dating back to the late 19th and early 20th centuries. It is considered a generation three university, with the universities established between 1857 and 1947 being considered generation one and two, respectively.