IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Use of Novel Ensemble Machine Learning Approach for Social Media Sentiment Analysis

Use of Novel Ensemble Machine Learning Approach for Social Media Sentiment Analysis
View Sample PDF
Author(s): Ishrat Nazeer (School of Computer Science and Engineering, Lovely Professional University, Jalandhar, India), Mamoon Rashid (School of Computer Science and Engineering, Lovely Professional University, Jalandhar, India), Sachin Kumar Gupta (School of Electronics and Communication Engineering, Shri Mata Vaishno Devi University, Jammu, India) and Abhishek Kumar (School of Computer Science and IT, Jain University, Bangalore, India)
Copyright: 2021
Pages: 13
Source title: Analyzing Global Social Media Consumption
Source Author(s)/Editor(s): Patrick Kanyi Wamuyu (United States International University – Africa, Kenya)
DOI: 10.4018/978-1-7998-4718-2.ch002

Purchase

View Use of Novel Ensemble Machine Learning Approach for Social Media Sentiment Analysis on the publisher's website for pricing and purchasing information.

Abstract

Twitter is a platform where people express their opinions and come with regular updates. At present, it has become a source for many organizations where data will be extracted and then later analyzed for sentiments. Many machine learning algorithms are available for twitter sentiment analysis which are used for automatically predicting the sentiment of tweets. However, there are challenges that hinder machine learning classifiers to achieve better results in terms of classification. In this chapter, the authors are proposing a novel feature generation technique to provide desired features for training model. Next, the novel ensemble classification system is proposed for identifying sentiment in tweets through weighted majority rule ensemble classifier, which utilizes several commonly used statistical models like naive Bayes, random forest, logistic regression, which are weighted according to their performance on historical data, where weights are chosen separately for each model.

Related Content

Stavroula Kalogeras. © 2021. 15 pages.
Ishrat Nazeer, Mamoon Rashid, Sachin Kumar Gupta, Abhishek Kumar. © 2021. 13 pages.
Marwa Mallouli Ben Zouitina, Zouhour Smaoui Hachicha. © 2021. 20 pages.
Esra Bozkanat. © 2021. 19 pages.
Yowei Kang, Kenneth C. C. Yang. © 2021. 20 pages.
Patrick Kanyi Wamuyu. © 2021. 33 pages.
José Duarte Santos, Steffen Mayer. © 2021. 17 pages.
Body Bottom