Adaptive technical specification template: graph-based model, structural analysis, and automated verification algorithm

In the context of increasing heterogeneity in software development practices and documentation standards, ensuring the completeness and structural consistency of technical specifications remains a complex and labor-intensive task. Existing regulatory frameworks, including GOST 34, IEEE 830, ISO/IEC/IEEE 29148, and the Volere methodology, propose different approaches to structuring requirements; however, their simultaneous use in real-world projects often results in section duplication, structural inconsistency, and significant manual verification effort. This paper proposes an adaptive technical specification template based on a parameterized graph model that enables the formal integration of a mandatory regulatory core with flexibly connected extensions depending on the software type, industry-specific requirements, and the required level of detail. An automated structural verification algorithm for DOCX and PDF documents is developed, combining hierarchy extraction with fuzzy heading matching. Template adaptability metric has been introduced. Experimental validation on real-world technical specifications demonstrates structural extraction accuracy of up to 92 % for DOCX documents. The proposed approach can serve as a basis for intelligent tools for analyzing technical documentation.

1. Frattini J. Identifying Relevant Factors of Requirements Quality: An Industrial Case Study. In: Requirements Engineering: Foundation for Software Quality: 30th International Working Conference, REFSQ 2024, 08–11 April 2024, Winterthur, Switzerland. Cham: Springer; 2024. P. 20–36. https://doi.org/10.1007/978-3-031-57327-9_2

2. AbuSalim S.W.G., Ibrahim R., Mostafa S.A., Wahab J.A. Analyzing the Impact of Assessing Requirements Specifications on the Software Development Life Cycle. In: Computational Science and Its Applications – ICCSA 2020: 20th International Conference: Proceedings: Part VI, 01–04 July 2024, Cagliari, Italy. Cham: Springer; 2020. P. 632–648. https://doi.org/10.1007/978-3-030-58817-5_46

3. Wei X., Wang Zh., Yang Sh. An Automatic Generation and Verification Method of Software Requirements Specification. Electronics. 2023;12(12). https://doi.org/10.3390/electronics12122734

4. Umar M.A., Lano K., Abubakar A.K. Automated requirements engineering framework for agile model-driven development. Frontiers in Computer Science. 2025;7. https://doi.org/10.3389/fcomp.2025.1537100

5. Weitl F., Jakšić M., Freitag B. Towards the Automated Verification of Semi-Structured Documents. Data & Knowledge Engineering. 2009;68(3):292–317. https://doi.org/10.1016/j.datak.2008.10.003

6. Moon S., Lee G., Chi S. Automated system for construction specification review using natural language processing. Advanced Engineering Informatics. 2022;51. https://doi.org/10.1016/j.aei.2021.101495

7. Nasyrov N., Komarov M., Tartynskikh P., Gorlushkina N. Automated formatting verification technique of paperwork based on the gradient boosting on decision trees. Procedia Computer Science. 2020;178:365–374. https://doi.org/10.1016/j.procs.2020.11.038

8. Ghosh Sh., Elenius D., Li W., et al. ARSENAL: Automatic Requirements Specification Extraction from Natural Language. In: NASA Formal Methods: 8th International Symposium, 07–09 June 2016, Minneapolis, MN, USA. Cham: Springer; 2016. P. 41–46. https://doi.org/10.1007/978-3-319-40648-0_4

9. Rodrigues A., da Silva A.R. Validation of Rigorous Requirements Specifications and Document Automation with the ITLingo RSL Language. arXiv. URL: https://doi.org/10.48550/arXiv.2312.10822 [Accessed 19th December 2025].

10. Jiang Sh., Hu J., Magee Ch.L., Luo J. Deep Learning for Technical Document Classification. IEEE Transactions on Engineering Management. 2022;71:1163–1179. https://doi.org/10.1109/TEM.2022.3152216

11. Baratov D., Aripov N., Jumanov X., Muhiddinov O. Automated technology of control of technical documentation of automation and telemechanics. AIP Conference Proceedings. 2023;2612(1). https://doi.org/10.1063/5.0135326

12. Biswas Ch., Das S. ARIA-QA: AI-agent based requirements inspection and analysis through question answering. Innovations in Systems and Software Engineering. 2024;21:1009–1024. https://doi.org/10.1007/s11334-024-00589-8

13. Kolahdouz Rahimi Sh., Lano K., Tehrani S., et al. Comparative evaluation of NLP approaches for requirements formalisation. In: Proceedings of the 12th International Conference on Model-Based Software and Systems Engineering, 21–23 February 2024, Rome, Italy. SCITEPRESS; 2024. P. 125–132. https://doi.org/10.5220/0012318700003645

14. Sonbol R., Rebdawi Gh., Ghneim N. The Use of NLP-Based Text Representation Techniques to Support Requirement Engineering Tasks: A Systematic Mapping Review. IEEE Access. 2022;10:62811–62830. https://doi.org/10.1109/ACCESS.2022.3182372

15. Ferrari A., Zhao L., Alhoshan W. NLP for Requirements Engineering: Tasks, Techniques, Tools, and Technologies. In: 2021 IEEE/ACM 43rd International Conference on Software Engineering: Companion Proceedings (ICSE-Companion), 25–28 May 2021, Madrid, ES. IEEE; 2021. P. 322–323. https://doi.org/10.1109/ICSE-Companion52605.2021.00137

16. Xu D., Chen W., Peng W., et al. Large language models for generative information extraction: a survey. Frontiers of Computer Science. 2024;18(6). https://doi.org/10.1007/s11704-024-40555-y