Fake News Detection on Social Media Using Ensemble Methods  

作　　者：Muhammad Ali Ilyas Abdul Rehman Assad Abbas Dongsun Kim Muhammad Tahir Naseem Nasro Min Allah 

机构地区：[1]Department of Computer Science,COMSATS University,Islamabad,45550,Pakistan [2]School of Computer Science and Engineering,Kyungpook National University,Daegu,41566,Republic of Korea [3]Department of Computer Science and Engineering,Korea University,Seoul,02841,Republic of Korea [4]Department of Electronic Engineering,Yeungnam University,Gyeongsan-si,38541,Republic of Korea [5]Department of Computer Science,College of Computer Science and Information Technology,Imam Abdulrahman Bin Faisal University,Dammam,34223,Saudi Arabia

出　　处：《Computers, Materials & Continua》2024年第12期4525-4549,共25页计算机、材料和连续体（英文）

基　　金：supported by the MSIT(Ministry of Science and ICT),Korea,under the ICT Creative Consilience Program(IITP-2024-2020-0-01819);supervised by the IITP(Institute for Information&Communications Technology Planning&Evaluation).

摘　　要：In an era dominated by information dissemination through various channels like newspapers,social media,radio,and television,the surge in content production,especially on social platforms,has amplified the challenge of distinguishing between truthful and deceptive information.Fake news,a prevalent issue,particularly on social media,complicates the assessment of news credibility.The pervasive spread of fake news not only misleads the public but also erodes trust in legitimate news sources,creating confusion and polarizing opinions.As the volume of information grows,individuals increasingly struggle to discern credible content from false narratives,leading to widespread misinformation and potentially harmful consequences.Despite numerous methodologies proposed for fake news detection,including knowledge-based,language-based,and machine-learning approaches,their efficacy often diminishes when confronted with high-dimensional datasets and data riddled with noise or inconsistencies.Our study addresses this challenge by evaluating the synergistic benefits of combining feature extraction and feature selection techniques in fake news detection.We employ multiple feature extraction methods,including Count Vectorizer,Bag of Words,Global Vectors for Word Representation(GloVe),Word to Vector(Word2Vec),and Term Frequency-Inverse Document Frequency(TF-IDF),alongside feature selection techniques such as Information Gain,Chi-Square,Principal Component Analysis(PCA),and Document Frequency.This comprehensive approach enhances the model’s ability to identify and analyze relevant features,leading to more accurate and effective fake news detection.Our findings highlight the importance of a multi-faceted approach,offering a significant improvement in model accuracy and reliability.Moreover,the study emphasizes the adaptability of the proposed ensemble model across diverse datasets,reinforcing its potential for broader application in real-world scenarios.We introduce a pioneering ensemble technique that leverages both machine-learning

关 键 词：Fake news detection Machine Learning(ML) Deep Learning(DL) CHI-SQUARE ensembling 

分 类 号：G206[文化科学—传播学]