A Comparative Study of Machine Learning Models for Detecting Fake News Content in Bahasa Indonesia Online Media

The rapid expansion of online media platforms and online channels has democratized information access and dissemination, including news content. However, the ease of creation and dissemination has also facilitated the spread of fake news. Machine learning offers promising solutions, yet research significantly relies on specific languages. This study investigates the detection of Indonesian-language fake news using popular classification algorithms (Decision Tree, Naive Bayes, Random Forest, K-Nearest Neighbors, Logistic Regression, and Support Vector Machine) on the “Indonesian Hoax News Detection” Dataset. We employ the Trigram Model for data characteristic extraction. K-Nearest Neighbor achieved the highest accuracy (72.8%) and precision (76.2%) compared to other models. While F1-Scores showed minimal variation between algorithms, the results indicate K-Nearest Neighbor’s potential for Indonesian fake news detection systems.