Bilingual fake-news detection in low-resource media: A Transformer-based framework for Nepali–English content

Abstract

The spread of misinformation in Nepal, especially across the bilingual Nepali–English media landscape, represents a significant threat to informed public discourse. Existing solutions for low-resource languages often rely on traditional machine learning or simple recurrent neural networks, which struggle with the morphological complexity of Nepali and the semantic nuances of code-switched content. This paper presents a robust, production-ready framework for fake news detection that transitions from traditional ensembles to a state-of-the-art transformer-based architecture. We fine-tuneXLM-RoBERTa (XLM-R), a multilingual model optimized for cross-lingual transfer, on a newly curated and balanced corpus of16,000 articles(8,000 real, 8,000 fake). Unlike “black-box” approaches, we integrateSHAP (SHapley Additive exPlanations)to provide word-level interpretability, allowing users to understandwhyan article is flagged. The model achieves superior performance with an accuracy of99.53%and an F1-score of0.995, significantly outperforming baseline Bi-LSTM ensembles. The system is deployed as a scalable web application and browser extension using a microservices architecture (Django/React) backed byPostgreSQL, ensuring high concurrency and data persistence. We further address the ethical implications of automated detection by implementing strict privacy protocols and bias mitigation strategies for sensitive political content.