Enhancing Technical Document Compliance Review through a Context-Aware Generative AI Framework
DOI:
https://doi.org/10.32628/CSEIT25113329Abstract
The rigorous review of technical documentation for compliance with linguistic, domain-specific, and document- type standards is a critical, yet often labor-intensive and error-prone process. This paper presents a novel Generative AI (GenAI) based system designed to automate and significantly enhance the accuracy of technical document compliance checking. Our framework leverages Transformer-based Generative AI models within a hybrid architecture that synergizes Retrieval-Augmented Generation (RAG), semantic rule interpretation, and deep contextual analysis derived from a document graph. The system integrates three distinct layers of rule enforcement: language grammar, domain/standards specific rules, and document type specific requirements. A continuous human-in-the-loop feedback mechanism, driven by Reinforcement Learning from Human Feedback (RLHF), ensures iterative model refinement and rule base adaptation. The system provides a comprehensive compliance score and actionable, granular, and context-aware review comments. This paper details the implemented methodologies and materials, focusing on the strategies employed for achieving high accuracy and reliability in automated technical document review.
📊 Article Downloads
References
P. N. Johnson-Laird, "The manual review bottleneck in technical documentation," Journal of Technical Writing and Communication, vol. 45, no. 3, pp. 259-278, 2015.
R. S. Systems and S. P. Review, "Automated rule- based checking of technical specifications: Limitations and challenges," IEEE Transactions on Automation Science and Engineering, vol. 10, no. 4, pp. 1081-1090, 2013.
J. Devlin, M. Chang, K. Lee, and K. Toutanova, "BERT: Pre-training of deep bidirectional transformers for language understanding," arXiv preprint arXiv:1810.04805, 2018.
P. Lewis, E. Perez, A. Piktus, F. Petroni, V. Karpukhin, N. Goyal, H. Küttler, M. Lewis, Y. Gao, D. Oguz, and A. Joulin, "Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks," Advances in Neural Information Processing Systems, vol. 33, pp. 9459-9474, 2020.
J. Gao, "Confidence calibration in neural networks," PhD diss., Carnegie Mellon University, 2017.
A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, ... & O. Polosukhin, "Attention is all you need," Advances in Neural Information Processing Systems, vol. 30, 2017.
L. Ouyang, J. Wu, X. Zhang, H. Jiang, L. Tarlow, T. Kritchevsky, ... & J. Schulman, "Training language models to follow instructions with human feedback," Advances in Neural Information Processing Systems, vol. 35, pp. 27730-27744, 2022.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 International Journal of Scientific Research in Computer Science, Engineering and Information Technology

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