Sentiment Analysis of Public Opinions about Universitas Royal Using the Support Vector Machine Method

Abstract

Instagram is one of the communication platforms used by Universitas Royal to interact with the public and serves as a space for users to express opinions, responses, and evaluations of the institution. The large volume of available comments presents challenges in identifying sentiment patterns consistently through manual analysis. This study aims to analyze public sentiment toward Universitas Royal based on Instagram comments using the Support Vector Machine (SVM) method. The dataset consists of 2,457 comments categorized into three sentiment classes: negative, neutral, and positive. The research stages include text preprocessing, feature weighting using Term Frequency–Inverse Document Frequency (TF-IDF), data splitting at an 80:20 ratio, model training, and evaluation using accuracy, precision, recall, and F1-score. The initial evaluation results show that the SVM model achieved 100% across all evaluation metrics on 492 test samples. These results indicate a very high level of agreement between the model predictions and the reference labels in the test dataset. However, the possibility of duplicate comments requires further examination to ensure the reliability of the evaluation and the model's generalization ability. This study has the potential to support the analysis of public opinions toward Universitas Royal as a basis for evaluating the institution's communication through Instagram.

.