Implementation of Support Vector Machine Method with TF-IDF for Sentiment Analysis of the Al-Zaytun Islamic Boarding School Controversy

Abstract

Al-Zaytun Islamic Boarding School in Indramayu, West Java, has attracted public attention on social media. The previous Eid prayer went viral because men and women stood in the duplicate prayer rows. In addition, several other aspects also drew public attention, such as the Friday prayer call style being different from the usual, introducing Jewish greetings, and allegedly allowing students to commit adultery, with the sin being redeemable for a certain amount of money. These controversies naturally sparked various reactions from the Indonesian public. This study employs the Support Vector Machine (SVM) method combined with Term Frequency-Inverse Document Frequency (TF-IDF) word weighting to evaluate public sentiment regarding various controversies associated with the Al-Zaytun Islamic boarding school. The data used in this research consists of tweets collected through a scraping process using Tweet Harvest with several relevant keywords. The results are analyzed to classify sentiment into three categories: positive, neutral, and hostile. The entire process is carried out systematically to obtain classification results that are both accurate and relevant to the ongoing social phenomena. Therefore, this study aims to implement the Support Vector Machine (SVM) algorithm to classify Twitter user sentiments towards the Al-Zaytun Islamic Boarding School controversy. The research collected 1,018 tweets through a scraping process using Tweet Harvest via Google Collab, with keywords such as "alzaytun," "zaytun," "panji gumilang," and "al-zaytun." The sentiment distribution consisted of 133 positive sentiments, 313 negative sentiments, and 572 neutral sentiments. Based on the classification evaluation results, the Support Vector Machine algorithm achieved an accuracy of 76%, a precision of 78.3%, a recall of 67.6%, and an F1 score of 69.6%.