Sentiment Analysis of people's opinions in E-Learning based on Support Vector Machine

Abstract

The rapid development and popularity of social media have allowed people to express their ideas and views, and feelings through social networks such as Twitter and Meta. This research focused on sentiment analysis of social media content about electronic learning (E-learning) since public opinions can help organizations and individuals. In this area, the social media content is abundant and unregulated. Therefore, sentiment analysis has become mainly an area of research interest. This study explores the machine learning approach for sentiment analysis of Twitter content to analyze people's opinions about E-learning. Using the Twitter API, the number of tweets collected was 42368. Tweets that involved E-learning were selected by using a programmable code written in Python. Then, a Support Vector Machine classifier was trained on the pre-processed data for analysing the tweets. The result of this study showed that the classifier's accuracy applied to the dataset was 92%. Generally, the people's opinions were positive toward E-learning.