<![CDATA[The corona virus pandemic that occurred in 2020 caused lecture activities to be carried out online to prevent the spread of the corona virus. Lectures that are held online receive a lot of opinion from the public and students. The large number of opinions regarding online lectures can be carried out by sentiment analysis to find out what opinions are expressed by the public and students. The process of sentiment analysis is carried out using the Multinomial Naïve Bayes method. The data used for the sentiment analysis process is 4,014 with the keywords "online lectures and online learning". The data will be cleaned first through a preprocessing process and then labeled using a text blob and categorized into positive, negative, and neutral classes. In this study, a comparison of accuracy results will also be carried out using the Information Gain feature selection in the hope of increasing accuracy results. Based on the test results that have been carried out, the Information Gain feature selection is proven to increase accuracy in the sentiment analysis process using Multinomial Naïve Bayes. The highest accuracy results using the Information Gain feature selection of 79.54%. While the highest accuracy results without using the Information Gain feature selection are 78.43%.]]>
Copyrights © 2023
