Hybrid Machine Learning Application in Classifying the Sentiment by Analyzing the Political Tweets


  • Reetu Jain


In this present techno-socio world, social media has dramatically changed the lives of peoples. Social media, especially Twitter, is vastly used for sharing information and thoughts, expressing opinions, seeking support and many other such activities. Scientists and strategists try to analyse the sentiments of the common people from their tweets.  By evaluating the sentiments of the common people, political scientists are capable of detecting early crises and the strategist could strategize the next move in favor of the political figures.  Because of this reason the comprehensive intention of the present study is to develop a robust classification algorithm that can be applied for sentiment analysis from the political tweets. In order to achieve this objective a hybrid machine learning classification algorithm is proposed in the study.  proposed hybridizes natural language processing (NLP) and long short term memory (LSTM). The NLP part of the model part of the proposed model extracts keywords that can rightly summarize the text of the tweets and correlate them with the positive or negative sentiments. On the other hand the LSTM part develops a predictive model that is capable of classifying the keywords into positive or negative sentiments. The potentiality of a proposed model lies in its applicability. Therefore the proposed model is applied to detect the sentiments of common people from their tweets. The Twitter data set comprises 1.6 million political tweets. The proposed model showed an accuracy of 78% and a loss of 0.456. The proposed model also showed a precision of 79% and F1 score of 0.78.




How to Cite

Jain, R. (2022). Hybrid Machine Learning Application in Classifying the Sentiment by Analyzing the Political Tweets. iJournals:International Journal of Software & Hardware Research in Engineering ISSN:2347-4890, 10(6). Retrieved from https://ijournals.in/journal/index.php/ijshre/article/view/134

Most read articles by the same author(s)

1 2 > >>