Exploring Sentiments: An In-Depth Analysis of Opinions in Education-Focused Tweets

Abstract

Investigating students' views on online learning is important for reinforcing the best educational offerings. Analyzing reviews serves as a sensible device to assess the numerous emotional polarities expressed by college students, losing mild at the acceptance or rejection of unique academic policies. Leveraging social media discussions proves valuable in delving into students' critiques even though the unstructured nature of these facts poses demanding situations and hinders prediction model performance. This research endeavours to address those demanding situations using Natural Language Processing (NLP) strategies for analyzing unstructured textual content. They look at specializing in amassing social media information from the Twitter database. To ensure the reliability of the findings, a robust method is applied to eliminate noisy, besides-the-point, and redundant information. An essential element influencing type accuracy is function selection. This study introduces a fuzzy C-means Algorithm choice-making model to streamline huge capabilities extracted from the text feature units. The overall performance of the polarity category for social media information is evaluated using accuracy, precision, consideration, and f-rating measures. The evaluation outcomes spotlight the prevalence of the fuzzy c means-based SVM (FIT2S-SVM) classifier, reaching a maximum accuracy rate of zero.97 and outperforming current methodologies.