BEYOND ALGORITHMS: AN INTEGRATED APPROACH TO FAKE NEWS DETECTION USING MACHINE LEARNING TECHNIQUES

Aug 4, 2026

Upload by Levi

The internet serves as a primary information channel while accelerating the spread of fake news that can shape public opinion and social decisions. Rather than relying on single-algorithm solutions with limited performance, this study compares machine learning models for fake news detection: SVC, XGBoost, and a Stacking Ensemble combining both. Text preprocessing uses IndoBERT to generate context-aware, semantically rich Indonesian representations. Results show the Stacking Ensemble achieves 82% accuracy, outperforming XGBoost (79%) and SVC (78%), supported by complementary strengths and improved text embeddings.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages14

Reading Time35 min

AccessFree to download

Introduction

Problem background and impact

Motivation for machine learning and NLP

Comparative models for detection

BEYOND ALGORITHMS: AN INTEGRATED APPROACH TO FAKE NEWS DETECTION USING MACHINE LEARNING TECHNIQUES Bimantyoso Hamdikatama1* Faculty of Communication and Informatics 1, Universitas Muhammadiyah Surakarta, Surakarta, Indonesia1 [www.ums.ac.id](www.ums.ac.id1)[1](www.ums.ac.id1) [bh972@ums.ac.id](bh972@ums.ac.id1)[1](bh972@ums.ac.id1)* (*) Corresponding Author (Responsible for the Quality of Paper Content) The creation is distributed under the Creative Commons Attribution-NonCommercial 4.0 International License. Abstract— The internet has become a major source of information, but it also facilitates the rapid spread of fake news, which can significantly influence public opinion and social decisions. While various techniques have been developed for detecting fake news, many studies focus on individual algorithms, which often result in suboptimal performance. This study addresses this gap by comparing machine learning models, including Support Vector Classification (SVC), XGBoost, and a Stacking Ensemble that combines both SVC and XGBoost, to determine the most effective approach for fake news detection. Text preprocessing was performed using IndoBERT, which provides context-aware and semantically rich text representations specifically for the Indonesian language. The evaluation results demonstrate that the Stacking Ensemble outperforms the individual models, achieving an accuracy of 82%, compared to 79% for XGBoost and 78% for SVC. This superior performance is attributed to the complementary strengths of the base models: SVC excels in handling highdimensional data, while XGBoost effectively manages imbalanced datasets and captures complex feature interactions. The use of IndoBERT further enhances model performance by improving text representation through contextual embeddings. These findings highlight the effectiveness of ensemble learning in enhancing predictive performance and robustness for fake news detection, demonstrating the potential of combining different machine learning techniques with advanced preprocessing methods to achieve more reliable results. Keywords: BERT, ensemble learning, SVC, XGBoost. Intisari—Internet telah menjadisumber informasi utama, tetapi juga memfasilitasi penyebaran berita palsusecara cepat, yang dapat memengaruhi opini publik dan pengambilan keputusan sosial secara signifikan. Meskipun berbagai teknik telah dikembangkan untuk mendeteksi berita palsu, banyak penelitianyang hanya berfokus pada algoritma individu, yang sering kali menghasilkan kinerja yang kurang optimal. Penelitian ini mengatasi kesenjangan tersebut dengan membandingkan model pembelajaran mesin, termasuk Support Vector Classification (SVC), XGBoost, dan Stacking Ensemble yang menggabungkan SVC dan XGBoost, untukmenentukan pendekatan paling efektif dalam mendeteksi berita palsu. Pemrosesan teks dilakukan menggunakan IndoBERT, yang menyediakan representasi teks yang kaya secara semantik dan kontekstual khusus untuk bahasa Indonesia. Hasil evaluasi menunjukkan bahwa Stacking Ensemble memiliki kinerja lebih baik dibandingkan model individu lainnya, dengan akurasi mencapai 82%, dibandingkan dengan 79% untuk XGBoost dan 78% untuk SVC. Kinerja unggul ini disebabkan oleh kekuatan komplementer dari model dasar: SVC unggul dalam menangani data berdimensi tinggi, sementara XGBoost efektif dalam mengelola dataset tidakseimbang dan menangkap interaksifituryang kompleks. Penggunaan IndoBERT semakin meningkatkan kinerja model dengan memperbaiki representasi teks melalui embedding kontekstual. Temuan ini menegaskan efektivitas pembelajaran ensemble dalam meningkatkan kinerja prediktif dan ketahanan sistem untuk deteksi berita palsu, serta menunjukkan potensi penggabungan berbagai teknik pembelajaran mesin dengan metode pra-pemrosesan lanjutan untuk mencapai hasilyang lebih andal. Kata Kunci: BERT, pembelajaran ensemble, SVC, XGBoost. INTRODUCTION In the digital era, social media and online platforms have transformed news

FAQs

Why is fake news detection challenging in online environments?

Fake news detection is difficult due to the large volume and wide variety of content circulating on platforms, which complicates reliable identification.

Which models are compared for fake news detection?

The study compares SVC, XGBoost, and a Stacking Ensemble that combines SVC and XGBoost.

How does IndoBERT improve model performance?

IndoBERT enhances performance by improving text representation through contextual embeddings that better capture semantic meaning in Indonesian.

Aug 4, 2026

Upload by Levi