Title: A deep learning model with effective tokenisation and feature extraction for detection of rumours in online social networks

Authors: Chandrakant Mallick; Sarojananda Mishra; Sumanjit Das; Bijay Kumar Paikaray

Addresses: Department of Computer Science and Engineering, Biju Patnaik University of Technology, Rourkela, Odisha, India ' Department of Computer Science and Engineering, Indira Gandhi Institute of Technology, Sarang, Odisha, India ' School of Electrical Sciences, Indian Institute of Technology, Bhubaneswar, Odisha, India ' Centre for Data Science, Department of Computer Science and Engineering, Siksha 'O' Anusandhan (Deemed to be University), Bhubaneswar, Odisha, India

Abstract: The proliferation of rumours on social media platforms, as well as their potential societal impact, presents a serious challenge that necessitates robust models for the precise detection and filtering of rumour posts in order to limit their harmful implications. This paper explores the persistent problem of rumour spread on social media platforms and its significant societal effects. In order to achieve highly accurate detection and filtration of rumour-related posts within online social networks, it introduces a novel deep learning model that makes use of long-short-term memory architecture along with reliable tokenisation and feature extraction techniques. The model's extraordinary effectiveness in identifying rumours is systematically assessed using well-known Twitter datasets and contrasted with other state-of-the-art models, demonstrating its efficacy in the detection of rumour posts.

Keywords: rumour detection; online social networks; deep learning; long-short-term memory; LSTM; tokenisation; feature extraction.

DOI: 10.1504/IJIMS.2025.146818

International Journal of Internet Manufacturing and Services, 2025 Vol.11 No.2, pp.93 - 113

Received: 06 Nov 2023
Accepted: 27 Jan 2024
Published online: 20 Jun 2025 *

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