The rapid expansion of digital platforms and social media has significantly accelerated the spread of fake news and misinformation, creating serious challenges to information integrity, public trust, and cyber security. Misleading information is often exploited for purposes such as social engineering, political manipulation, and information warfare, making timely and accurate detection increasingly important. Traditional manual verification methods are insufficient to manage the vast volume, speed, and complexity of online content generated every day. This project proposes an automated fake news and misinformation detection system that leverages Natural Language Processing (NLP) techniques along with Large Language Models (LLMs) to analyze and classify news content. The system performs text preprocessing and feature extraction to identify linguistic patterns, contextual signals, and semantic indicators that differentiate genuine news from fabricated or misleading information.

This paper presented TruthLens, a hybrid fake news and misinformation detection platform that combines a deterministic six-stage NLP pipeline with LLM-driven explainability using Llama 3 via Ollama. The system requires no labelled training data, no GPU for its deterministic stages, and no external cloud APIs, making it suitable for privacy-sensitive deployments. Experimental Evaluation on a 120-article benchmark achieves a weighted F1 score of 0.82.The proposed system contributes to enhancing information integrity by providing users with transparent, step-by-step reasoning behind every verdict, crossverification against six trusted news outlets, and actionable recommendations for further fact-checking. The system demonstrates an effective approach to tackling the growing challenge of misinformation by combining deterministic NLP-based analysis with LLM-driven explainability. By integrating structured content analysis with transparent reasoning, the system not only provides a credibility verdict but also helps users understand the underlying factors influencing the decision. The platform is designed with a focus on scalability, privacy, and usability, as it avoids dependency on external cloud services and operates efficiently without requiring high computational resources such as GPUs for its core processing stages. This makes the system suitable for deployment in resource-constrained and privacy-sensitive environments. Furthermore, the inclusion of trusted source verification enhances the reliability of the analysis by cross-checking news content against established media outlets. The system’s ability to generate interpretable AI explanations ensures that users are not only informed about the result but are also empowered to make their own judgments. In future work, the system can be extended by incorporating real-time data integration, multilingual support, and adaptive learning mechanisms to further improve accuracy and applicability across diverse information ecosystems. Additionally, integrating user feedback loops could help refine the scoring mechanism and enhance system performance over time.Overall, TruthLens contributes to strengthening digital information integrity, combating misinformation, and supporting informed decision-making in modern online environments.The system reduces the limitations of manual verification and enables faster identification of misinformation, contributing to improved cyber security and information integrity. Overall, TruthLens provides a scalable and practical approach to support reliableinformation consumption in digital platforms

[1]C. Castillo, M. Mendoza, and B. Poblete, ―Information credibility on Twitter,‖ in Proc. WWW, 2011, pp. 675–684. [2]M. Potthast, J. Köbabe, P. Strickson et al., ―A stylometric inquiry into Hyperpartisan and fake news,‖ in Proc. NAACL, 2018, pp. 231–240. [3]Y. Wang, ―LIAR, LIAR pants on fire: A new benchmark dataset for Fake news detection,‖ in Proc. ACL, 2017, pp. 422–426. [4]J. Devlin, M.-W. Chang, K. Lee, and K. Toutanova, ―BERT: Pretraining of deep bidirectional transformers for language understanding,‖ in Proc. NAACL, 2019, pp. 4171–4186. [5]K. Shu, D. Mahudeswaran, S. Wang, D. Lee, and H. Liu, ―FakeNewsNet: A data repository with news content, social context and dynamic information for studying fake news on social media,‖ Big Data, vol. 8, no. 3, pp. 171–188, 2020. [6] W. Guo, J. Chen, and M. Gao, ―Towards LLM-based automated factchecking with claim decomposition and evidence retrieval,‖ arXiv preprint arXiv:2310.15985, 2023. [7]L. Pan, W. Chen, H. Wang, and M. Kan, ―Fact- checking complex claims with program-guided reasoning,‖ in Proc. ACL, 2023, pp. 6981–6998. [8]X. Zhou and R. Zafarani, ―A survey of fake news: Fundamental Theories, detection methods, and opportunities,‖ ACM Computing Surveys, vol. 53, no. 5, pp. 1–40, 2021. [9]K. Popat, S. Mukherjee, A. Yates, and G. Weikum, ―DeClarE: Debunking fake news and false claims using evidence-aware deep Learning,‖ in Proc. EMNLP, 2018, pp. 22–32. J. Thorne, A. Vlachos, C. Christodoulopoulos, and A. Mittal, ―FEVER: A large-scale dataset for fact extraction and VERification,‖ in Proc. NAACL, 2018, pp. 809–819.

© 2025 International Journal of Engineering and Techniques (IJET).

Digital Object Identifier (DOI)

IJET assigns a unique DOI to every accepted article via Zenodo for persistent linking and citation. Your article’s DOI: https://doi.org/{{doi}}

Open DOI About Zenodo DOI

Close