Int J Performability Eng ›› 2022, Vol. 18 ›› Issue (6): 434-443.doi: 10.23940/ijpe.22.06.p6.434-443

Previous Articles     Next Articles

Hybrid Metaheuristic Approach for Detection of Fake News on Social Media

Poonam Naranga,*, Ajay Vikram Singha, and Himanshu Mongab   

  1. aAmity University, Noida, 201301, India;
    bJawaharlal Nehru Government Engineering College, Sundernagar, 175018, India
  • Submitted on ; Revised on ; Accepted on
  • Contact: * E-mail address: poonam.ahuja1@s.amity.edu

Abstract: The modern environment has faced numerous issues due to fake news that machines or humans usually generate. It causes severe impacts in both political and social stages; therefore, each individual may face the effects of fake news. Thus, recent research mainly focuses on enabling a highly effective detection system to mitigate the negative influence of social media. A novel framework is developed to detect and classify fake news from social media automatically. In the proposed model, an Apache Spark technique is incorporated with the deep hybrid learning based on improved CNN with hybrid Black Widow Optimization (BWO) algorithm and Mothfly Optimization algorithm (MOA) (HM-BWO) and LSTM. This framework provides better classification and detection of fake news from social media environments. The proposed model is implemented into the platform of MATLAB, and its performance is analyzed through the performance metrics, including accuracy, loss, precision, F1-measure, and recall.

Key words: natural language processing, Apache Spark, BWO-Mothfly optimization, CNN-LSTM, deep learning, machine learning