Abstract :

The Rapid Proliferation Of Digital Media Platforms Has Significantly Increased The Spread Of Misinformation And Fake News, Posing Serious Threats To Public Trust, Democratic Processes, And Societal Stability. Traditional Methods Of Detecting Fake News Often Rely On Either Manual Verification Or Single-model Automated Approaches, Which Are Limited In Scalability, Accuracy, And Contextual Understanding. To Address These Challenges, This Project Introduces TruthGuard AI, A Hybrid Fake News Detection Platform That Integrates Rule-based Linguistic Analysis, Machine Learning Inference Using TF-IDF, And A Large Language Model (LLM) Verification Layer.The Proposed System Operates In A Multi-stage Pipeline Designed To Enhance Both Efficiency And Reliability. In The First Stage, Rule-based Analysis Evaluates Textual Content For Clickbait Patterns, Sensational Language, And Known Fake-news Indicators Such As Exaggerated Claims Or Conspiracy-related Keywords. This Stage Provides A Quick Heuristic Filter That Identifies Suspicious Linguistic Features. In The Second Stage, The System Applies A TF-IDF-based Machine Learning Model To Analyze Term Importance And Contextual Relevance, Enabling Statistical Classification Of News As Credible Or Fake. This Approach Improves Detection Accuracy By Leveraging Historical Patterns In Labeled Datasets.The Third Stage Introduces An Advanced Verification Layer Powered By A Local LLM (via Ollama), Which Performs Deep Contextual Reasoning. The LLM Analyzes The Content For Bias, Missing Sources, Logical Inconsistencies, And Credibility Signals, Offering An Explainable And Human-like Interpretation Of The News. This Hybrid Architecture Ensures That The Limitations Of One Method Are Compensated By The Strengths Of Others, Resulting In A More Robust Detection System.The Platform Is Implemented As A User-friendly Desktop Application Using Python’s Tkinter And Ttkbootstrap Libraries, Providing An Interactive Interface For Users To Input News Articles Or Headlines. The System Outputs A Reliability Score, A Verdict (credible, Suspicious, Or Fake), And A Detailed Reasoning Log That Enhances Transparency And User Trust.Experimental Results Demonstrate That Combining Heuristic, Statistical, And Deep Learning-based Approaches Significantly Improves Detection Performance Compared To Standalone Methods. The System Also Supports Scalability And Extensibility, Allowing Integration With Real-time News Feeds And Cloud-based LLM Services In Future Enhancements.In Conclusion, TruthGuard AI Represents A Comprehensive And Practical Solution For Combating Misinformation. By Combining Traditional And Modern AI Techniques, It Not Only Detects Fake News Effectively But Also Provides Explainable Insights, Making It Suitable For Real-world Deployment In Journalism, Education, And Social Media Monitoring.

Published:

07-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

1427-1436

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite