Title: Enhancing information integrity on social media: a deep learning approach to fake news classification using LSTM and GloVe

Authors: Bhabesh Ranjan Kar; Bijay Kumar Paikaray; Chandrakant Mallick

Addresses: Centre for Data Science, Department of Computer Science and Engineering, Siksha 'O' Anusandhan (Deemed to be) University, Bhubaneswar, Odisha, 751013, India ' Centre for Data Science, Department of Computer Science and Engineering, Siksha 'O' Anusandhan (Deemed to be) University, Bhubaneswar, Odisha, 751013, India ' Department of Computer Science and Information Technology, Gandhi Institute of Technological Advancement (GITA) Autonomous College, Bhubaneswar, Odisha, 752054, India

Abstract: With the growth of digital platforms, it has become crucial to identify fake news early to alert and protect individuals from its harmful effects. To deal with this problem, detecting fake news and understanding how it spreads are important for users on social media platforms. This work uses deep learning and advanced natural language processing (NLP) methods to classify real and fake news. The suggested model employs a long short-term memory (LSTM) neural network in combination with global vectors for word representations (GloVe) for text vectorisation and employs tokenisation for feature extraction to enhance its performance. This approach yields remarkable outcomes, attaining an accuracy rate of 98.15%. This study provides an efficient method for identifying fake news, reducing the spread of false information, and promoting informed decisions for users on social media platforms.

Keywords: deep learning; LSTM; long short-term memory; word embedding; tokenisation; fake news.

DOI: 10.1504/IJISCM.2025.148634

International Journal of Information Systems and Change Management, 2025 Vol.15 No.2, pp.206 - 227

Received: 25 Jan 2025
Accepted: 25 May 2025
Published online: 16 Sep 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article