Analysis of MyASN Application User Sentiment on Twitter Using the Support Vector Machine Algorithm

Abstract

This study aims to analyze user sentiment toward the MyASN application based on public discussions on X. A total of 1059 tweets were collected through a web scraping process using the Tweet-Harvest tool on Google Colab, with data obtained from tweets containing keywords and hashtags related to MyASN and BKN. The collected data were preprocessed through case folding, tokenizing, normalization, stopword removal, and stemming, then represented using the TF-IDF weighting scheme. Sentiment labels were assigned into three categories: positive, negative, and neutral. The classification process employed the Support Vector Machine (SVM) algorithm, with data divided into 80% training data and 20% testing data. The experimental results show that the Support Vector Machine (SVM) algorithm achieved an accuracy rate of 98.1% with a precision value of 0.983, a recall of 0.981, and an F1-score of 0.982. Evaluation based on sentiment class shows that in the negative class, SVM produced a precision of 1.000, a recall of 0.977, and an F1-score of 0.989. In the neutral class, it achieved a precision of 0.929, a recall of 1.000, and an F1-score of 0.963, while in the positive class, SVM achieved a precision of 0.885, a recall of 1.000, and an F1-score of 0.939. These results show that the indicates that the implemented SVM model demonstrates strong reliability in handling text-based sentiment classification, particularly in datasets with imbalanced sentiment distributions. Overall, the results demonstrate that SVM is effective in capturing user sentiment patterns and can provide meaningful insights for evaluating and improving the MyASN application service.