Communication Sentiment Analyzer using Machine Learning with Naive Bayes Bernoullinb
M. Prabu1, Mayank Singh Aithani2, Niroj Deb3, Pratyush Joshi4
1Mr. Prabu, Asst. Professor Srm Institute of Science and Technology, Ramapuram, Chennai. India.
2Mayank Singh Aithani, B. Tech Student at SRM Institute of Science and Technology, Ramapuram Chennai, India.
3Niroj Deb, B. Tech Student at SRM Institute of Science and Technology, Ramapuram Chennai, India.
4Pratyush Joshi, B. Tech Student at SRM Institute of Science and Technology, Ramapuram Chennai, India.
Manuscript received on September 22, 2019. | Revised Manuscript received on October 30, 2019. | Manuscript published on October 30, 2019. | PP: 5976-5979 | Volume-9 Issue-1, October 2019 | Retrieval Number: A1610109119/2019©BEIESP | DOI: 10.35940/ijeat.A1610.109119
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In this never-ending social media era it is estimated that over 5 billion people use smartphones. Out of these, there are over 1.5 billion active users in the world. In which we all are a major part and before opening our messages we all are curious about what message we have received. No doubt, we all always hope for a good message to be received. So Sentiment analysis on social media data has been seen by many as an effective tool to monitor user preferences and inclination. Finally, we propose a scalable machine learning model to analyze the polarity of a communicative text using Naive Bayes’ Bernoulli classifier. This paper works on only two polarities that is whether the sentence is positive or negative. Bernoulli classifier is used in this paper because it is best suited for binary inputs which in turn enhances the accuracy of up to 97%.
Keywords: Sentiment analysis, Machine learning, Polarity, Naive Bayes’ Bernoulli.