| _version_ | 1866902076093628416 |
|---|---|
| author | Hamrouni, Hassen |
| author_facet | Hamrouni, Hassen |
| contents | <p>This paper presents Trust Sense, an explainable artificial intelligence system designed to detect psychological manipulation in digital content, including text, audio, and video.</p> <p>The proposed approach integrates natural language processing, multimodal analysis, and cognitive psychology principles to evaluate content credibility beyond traditional fake news detection methods. The system introduces a novel Manipulation Index (MI) and a Trust Score (TS), combining emotional intensity, urgency signals, bias indicators, and persuasive patterns.</p> <p>The architecture is based on a modular and scalable design, supporting real-time analysis and future integration into automated media pipelines such as AI-generated news broadcasting.</p> <p>Experimental results demonstrate strong performance across multiple evaluation metrics, highlighting the system’s ability to identify manipulative patterns and provide interpretable outputs.</p> <p>This work contributes to the field of explainable AI (XAI) by bridging technical analysis with psychological interpretation, offering a new perspective on misinformation detection.</p> <p>This paper is a preprint and has not yet undergone peer review.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_19616412 |
| institution | Zenodo |
| language | eng |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Trust Sense: An Explainable AI System for Detecting Psychological Manipulation in News Content Hamrouni, Hassen <p>This paper presents Trust Sense, an explainable artificial intelligence system designed to detect psychological manipulation in digital content, including text, audio, and video.</p> <p>The proposed approach integrates natural language processing, multimodal analysis, and cognitive psychology principles to evaluate content credibility beyond traditional fake news detection methods. The system introduces a novel Manipulation Index (MI) and a Trust Score (TS), combining emotional intensity, urgency signals, bias indicators, and persuasive patterns.</p> <p>The architecture is based on a modular and scalable design, supporting real-time analysis and future integration into automated media pipelines such as AI-generated news broadcasting.</p> <p>Experimental results demonstrate strong performance across multiple evaluation metrics, highlighting the system’s ability to identify manipulative patterns and provide interpretable outputs.</p> <p>This work contributes to the field of explainable AI (XAI) by bridging technical analysis with psychological interpretation, offering a new perspective on misinformation detection.</p> <p>This paper is a preprint and has not yet undergone peer review.</p> |
| title | Trust Sense: An Explainable AI System for Detecting Psychological Manipulation in News Content |
| url | https://doi.org/10.5281/zenodo.19616412 |