Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM68757
Volume:
10 |
11 |
Nov 2026
Pages:
1-6
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.

Master of Computer Application
Jaihind Institute of Management and Research, Kuran, Pune
Master of Computer Application
Jaihind Institute of Management and Research, Kuran, Pune
The exponential proliferation of unverified digital narratives across decentralized social media and syndication platforms poses severe systemic threats to public discourse, democratic integrity, and institutional trust. Deceptive articles are crafted to exploit cognitive heuristics and confirmation biases, exhibiting subtle stylometric, syntactic, and semantic manipulation that evades manual forensic auditing. In this investigation, we introduce an end-to-end scalable computational framework for automated fake news identification utilizing advanced machine learning and natural language processing (NLP). We establish a multi-tier feature representation paradigm combining n-gram term frequency-inverse document frequency (TF-IDF) smoothing, stylometric readability markers, and dense contextual transformer encodings. Comprehensive benchmarking is conducted across 72,134 empirical articles derived from the WELFake and ISOT standardized corpora. Experimental evaluations reveal that the online Passive-Aggressive classifier achieves a remarkable 94.8% classification accuracy and 94.6% macro F1-score with an ultra-low inference latency of 3.1 ms per document, establishing superior suitability for high-throughput edge stream ingestion. Furthermore, an ensemble stacking classifier incorporating bidirectional self-attention achieves 98.4% ROC-AUC, outperforming conventional probabilistic baselines. Model decisions are rendered transparent through Shapley additive exPlanations (SHAP), providing verifiable, interpretable attributions for real-time content moderation.
KEYWORDS
This investigation validates the application of advanced machine learning and NLP feature engineering for robust, high-velocity fake news detection. Through systematic evaluations across 72,134 empirical news articles, we demonstrated that combining sub-linear TF-IDF n-grams with stylometric syntactic markers delivers highly competitive classification fidelity while preserving real-time inference latency. The Passive-Aggressive classifier proved exceptionally effective for high-throughput edge deployment, attaining 94.8% accuracy and 3.1 ms inference latency. For forensic offline auditing, transformer-based ensemble architectures provide state-of-the-art discrimination, achieving 98.4% ROC-AUC. Integrating SHAP explainability bridges the gap between algorithmic predictions and human editorial oversight, illuminating deceptive tokens with mathematical rigor. Future extensions will investigate cross-lingual transfer learning, multimodal fusion incorporating computer vision forensics for synthetic image manipulation, and decentralized knowledge graph verification to combat automated generative disinformation campaigns.
[1] S. Vosoughi, D. Roy, and S. Aral, "The spread of true and false news online," Science, vol. 359, no. 6380, pp. 1146-1151, 2018. [2] K. Shu, A. Sliva, S. Wang, J. Tang, and H. Liu, "Fake news detection on social media: A data mining perspective," ACM SIGKDD Explorations, vol. 19, no. 1, pp. 22-36, 2017. [3] H. Ahmed, I. Traore, and S. Saad, "Detecting opinion spams and fake news using text classification," Security and Privacy, vol. 1, no. 1, p. e9, 2018. [4] P. K. Verma, P. Agrawal, I. Amorim, and R. Prodan, "WELFake: Word Embedding Over Linguistic Features for Fake News Detection," IEEE Transactions on Computational Social Systems, vol. 8, no. 4, pp. 881-893, 2021. [5] K. Crammer, O. Dekel, J. Keshet, S. Shalev-Shwartz, and Y. Singer, "Online passive-aggressive algorithms," Journal of Machine Learning Research, vol. 7, pp. 551-585, 2006. [6] J. Devlin, M. Chang, K. Lee, and K. Toutanova, "BERT: Pre-training of deep bidirectional transformers for language understanding," in Proc. NAACL-HLT, 2019, pp. 4171-4186. [7] Y. Liu et al., "RoBERTa: A robustly optimized BERT pretraining approach," arXiv preprint arXiv:1907.11692, 2019. [8] S. M. Lundberg and S.-I. Lee, "A unified approach to interpreting model predictions," in Advances in Neural Information Processing Systems (NeurIPS), 2017, pp. 4765-4774. [9] 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, 2020. [10] V. Perez-Rosas, B. Kleinberg, A. Lefevre, and R. Mihalcea, "Automatic detection of fake news," in Proc. 27th International Conference on Computational Linguistics (COLING), 2018, pp. 3391-3401.
Publication Date
11/5/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
5
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...