A Blockchain Based Intelligent System for Urdu Information Veracity Assessment

Abstract

The increasing popularity of digital and online news has raised serious apprehensions about the socio-political impact of the spread of information, the decline of press freedom, and the need for stronger news-evaluation mechanisms. Fake news can attract a large audience and be very profitable, so it should be thoroughly vetted before it is disseminated to the public. Biased news, often driven by profit motives and socio-political influences, contributes to mass manipulation and societal ambiguity. The proposed research work presents a model to evaluate the fake news in Urdu language using machine learning methods build upon on the security of blockchain smart contract. The system aims to improve Urdu news accuracy by the integration of machine learning classification with human professional views built on the top of decentralized platform. To train NLP (natural language processing) model the data has continuously fed, achieving significant accuracy improvements from 87% to 90% over nine training rounds. The blockchain implementation on the remix IDE test network has been done. The simulation shows the efficacious application of this proposed research work.