Hybrid System for Verifying the Credibility of Information (Rules-based and AI-based)

Please use these citation keys if you use the code

Fact Checking System: Information Credibility Verification

Hybrid system combining predicate logic rules with machine learning for automated fact checking and information credibility assessment.

Overview

This project implements a neuro-symbolic AI system that combines:

• Predicate Logic: Rule-based knowledge representation and reasoning
• Machine Learning: Neural networks for pattern recognition
• OWL Ontologies: Formal knowledge modeling with RDF/Turtle

The system assesses information source credibility, detects misinformation, and provides explainable verdicts on claim veracity.

Features

• Hybrid predicate logic + ML architecture
• OWL 2 DL ontology for domain modeling
• Rule-based fact verification engine
• Neural classification of credibility indicators
• Sentiment analysis and bias detection
• Source reputation scoring
• Automatic credibility classification (High/Medium/Low)
• Explainable AI (XAI) for transparency

Installation

pip install rdflib owlready2 scikit-learn torch nltk pandas numpy