Explainable AI for Classification and Prioritization of QHSE Audit Findings

November 2026
Vol-12, Issue-6
Paper ID: 29033
ISSN: 2395-4396
Downloads: 33
Download PDF

Abstract & Details

Research Area
Artificial Intelligence, Quality Management, QHSE Auditing, Risk Management and Engineering
Keywords
Explainable Artificial Intelligence Natural Language Processing QHSE Audits Risk Prioritization Human-in-the-Loop Review
Abstract
Quality, Health, Safety, and Environment (QHSE) audits generate narrative findings that are difficult to compare consistently across projects, functions, and audit cycles. This study proposes an explainable artificial intelligence (AI) decision-support framework for classifying these findings and prioritizing their review while preserving professional auditor judgment. The conceptual architecture combines natural language processing (NLP) for suggested QHSE domain labels with a separate, transparent Priority Risk Index (PRI) based on potential severity, recurrence, control weakness, regulatory exposure, and cross-functional reach. Separating classification confidence from risk priority is the central design choice: uncertainty about a finding's category must not automatically reduce its urgency. The framework also specifies local explanations for suggested classifications, a breakdown of the weighted risk factors, and mandatory human-review gates for elevated risk, low classification confidence, regulatory concerns, or findings spanning multiple domains. To illustrate the proposed scoring logic, six synthetic audit-style examples are presented without using confidential organizational records. These examples demonstrate how the index can rank hypothetical findings and how a reviewer could inspect the factors behind a suggested priority. They do not establish classifier accuracy, operational effectiveness, or generalizability. The paper outlines an empirical validation protocol using independently labeled and de-identified audit findings, comparison of classical and transformer-based NLP baselines, per-class performance measures, probability calibration, and auditor evaluation of explanation usefulness. Data protection, model versioning, human overrides, and monitoring for drift are also incorporated as governance requirements. The contribution is therefore a documented, auditable framework and analytical illustration rather than a trained or deployed AI system. Future research must test the method on authorized real-world data and assess whether its recommendations improve review consistency and timeliness without introducing unacceptable safety or compliance risks.

Author Information

# Name Institute / Affiliation
1 Dr. Feroz Khan Ismail Independent Researcher

How to Cite

Use the following formats to cite this article in your research.

APA Style
Ismail, Dr. Feroz Khan (2026). Explainable AI for Classification and Prioritization of QHSE Audit Findings. International Journal of Advance Research and Innovative Ideas In Education, 12(6), 30-36.
MLA Style
Ismail, Dr. Feroz Khan. "Explainable AI for Classification and Prioritization of QHSE Audit Findings." International Journal of Advance Research and Innovative Ideas In Education, vol. 12, no. 6, 2026, pp. 30-36.
IEEE Style
Dr. Feroz Khan Ismail, "Explainable AI for Classification and Prioritization of QHSE Audit Findings," International Journal of Advance Research and Innovative Ideas In Education, vol. 12, no. 6, pp. 30-36, 2026.
Vancouver Style
Ismail Dr. Feroz Khan. Explainable AI for Classification and Prioritization of QHSE Audit Findings. International Journal of Advance Research and Innovative Ideas In Education. 2026;12(6):30-36.
Harvard Style
Ismail, Dr. Feroz Khan (2026) 'Explainable AI for Classification and Prioritization of QHSE Audit Findings', International Journal of Advance Research and Innovative Ideas In Education, 12(6), pp. 30-36.
Chicago Style
Ismail, Dr. Feroz Khan. "Explainable AI for Classification and Prioritization of QHSE Audit Findings." International Journal of Advance Research and Innovative Ideas In Education 12, no. 6 (2026): 30-36.
Turabian Style
Ismail, Dr. Feroz Khan. "Explainable AI for Classification and Prioritization of QHSE Audit Findings." International Journal of Advance Research and Innovative Ideas In Education 12, no. 6 (2026): 30-36.

Export Citation

Related Research

Precision Agriculture Using Computer Vision: A Review of In-Field Disease Detection, Foliar Severity Estimation
Ashlesha Adsul et al. 2026 Precision Agriculture, Computer Vision, AI, Machine Learning
Download PDF
Development of a Custom Handwritten Kannada Dataset and CNN-BiLSTM-CTC Based OCR System
Nandini G R et al. 2026 OCR, Machine learning, Image Processing, CNN, Deep learning
Download PDF
CodeCoach AI: A Review of Automated Programming Assessment, Explainable Feedback, and an AI-Assisted Coding Practice Framework
Mr. Pratham Tembhurne et al. 2026 Artificial Intelligence in Education, focusing on automated code assessment and personalized feedback to improve programming skills.
Download PDF
PsyForm AI -An AI-Assisted Biopsychosocial Case Formulation Learning Framework.
DIKSHA BHASKAR BARSAGADE et al. 2026 Artificial Intelligence in Psychology and Medical Education
Download PDF
FutureEye: AI-Based Student Skill Gap Analysis and an Android Framework for Technical Assessment and Communication Practice
HARDIK BAWANKULE et al. 2026 AI-Based Skill Assessment and Personalized Learning using Machine Learning, LLMs, and Speech Technologies
Download PDF