Sentiment Analysis Using Lexical Analysis and Machine Learning-Based Approaches
Sunita Beniwal, Ashwani Kumar, Sunil Kumar, Mamta Sharma · 2026
Sentiment analysis is studying and analyzing the opinion, views, emotions, etc. of people on products, services, or topics. Movie reviews, product reviews, tweets, comments, etc. can be taken as data, and user&s;s emotions can be analyzed whether positive, negative, or neutral. Sentiment analysis can be done using many techniques and on different levels. Depending on the need whole sentence or document can be analyzed. Organizations can use the results of sentiment analysis for improving the quality of their products or services and sentiment analysis can aid in decision making. This paper discusses binary Sentiment Analysis (SA) i.e. positive or negative of YouTube comments on COVID-19 pandemic and IMBD movie reviews by using four different methods viz. Boolean Feature Multinomial Naïve Bayes (BMNB), Support Vector Machine (SVM), Lexical Analysis (LA), and Sentimentr Package (SP) of R language and also comparing the results. Our analysis outlines that BMNB and SVM performed better than others and the performance of these two was nearly the same with SVM leading by a small margin