Leveraging Emoticon Scores for Improved Sentiment Detection in Social Media Using Machine Learning Techniques

Social media has become a common platform for people to share their thoughts about products and individuals. There are different types of social media to analyzing specific sentiments in natural language is a challenging undertaking how to express their views, thoughts and emotions. These days, emoticons are commonly utilized and have become essential to written communication. The review frames a deliberate system for sentiment analysis, including information assortment, cleaning, order, and model preparation and testing. It utilizes freely accessible film audit datasets to analyse the circulation of positive and negative criticism across various dataset sizes. The centre of the exploration is information analysis, assessing different classifiers including Naïve Bayes, Multinomial Naïve Bayes, Logistic Regression, Bernoulli Naïve Bayes, Direct SVM/Straight SVC, and Nu SVM/Nu SVC. Calculated the emoticon on different datasets relapse arises as a hearty and powerful classifier across dataset sizes logistic regression techniques gives the best result in comparison to other techniques, while Direct SVM/Straight SVC shows a reliable expansion in exactness rates as dataset size develops.