Sentiment Analysis of Public Trust Towards Islamic Boarding School on Social Media Using Machine Learning Method

This study aims to analyse public sentiment towards Islamic boarding schools on social media using a machine learning approach. A total of 1,905 cleaned comments were collected from two platforms, Twitter (X) and YouTube, and then processed through the CRISP-DM stages, which include business understanding, data preparation, modelling, evaluation, and deployment. Pre-processing steps such as tokenisation, stemming, and labelling were applied to prepare the text data for analysis. The sentiment classification was carried out using five machine learning algorithms: Naïve Bayes, Decision Tree, Neural Network, Support Vector Machine (SVM), and Random Forest. The evaluation results revealed that Random Forest outperformed other models, achieving the highest accuracy (79%), F1-score (79%), precision (80%), and recall (79%), indicating a strong balance in identifying sentiment classes accurately and consistently. Additionally, the research implemented interactive visualisations using Streamlit, enabling the public and stakeholders to understand sentiment trends in a clear, data-driven format. These findings are expected to serve as a strategic foundation for Islamic boarding schools in building a positive image in the digital space and for further development of AI-based opinion monitoring systems.