Automated management systems in education: a comparative review and prospects for improvement

  1. Home
  2. Archive
  3. № 2 (71)
Author(s) Collection number Pages Download abstract Download full text
Трач О. Р., Трішин Ф. А. № 2 (71) 94-106 Image Image

Understanding and ensuring the quality of software (SW) specifications (QS) is a critical preventive measure that determines the reliability and validity of subsequent product quality metrics. This study synthesizes current data, confirming that the quality of the final SW is a direct function of the quality of the input specifications. Historical data shows that incomplete and changing requirements are the dominant causes of project failure, shifting the strategic focus from code metrics to requirements metrics as the primary risk factor.

An analysis of key standards such as ISO 25010 (SQuaRE) and IEEE 1061 de­monst­rates industry and academic consensus on the need for systematic measurement of requirements attributes such as completeness, consistency, and unambiguity. Among these attributes, ambiguity has the largest share of defect subtypes, highlighting its cri­tical impact on interpretation and implementation.

It has been found that the Goal/Question/Metric (GQM) methodology applied to requirements management (RM) practices in the CMMI model provides a structured framework for empirical validation of these metrics. The central practical indicator of specification quality is the Requirements Stability Ratio, whose high value directly correlates with unproductive rework and low functionality scalability.

The report also discusses the contradictions in the application of formal methods, which, although they eliminate ambiguity, can increase complexity and lead to “bugs in the specification itself,” which can account for up to 60% of all input defects. Significant limitations are associated with natural language quantification (NLP), including problems of lexical and contextual ambiguity, as well as dependence on the quality of training data.

However, the integration of large language models (LLMs) opens up promising avenues for automated defect detection and tracing, reducing post-release defects by up to 40% through early automated validation. The conclusions drawn emphasize a strategic shift in focus to requirements metrics as a key lever for increasing Return on Investment (ROI) in software development.

Keywords: software specification, requirements attributes, quality metrics, requirements metrics, software quality standards, development cost-effectiveness.

doi: 10.32403/1998-6912-2025-2-71-82-93


  • 1. F. N. Colakoglu et al. (2021) Software Product Quality Metrics: A Systematic Mapping Study. Retrieved October 17, 2025, from https://scispace.com/pdf/software-product-quality-metrics-a-systematic-mapping-study-2w1ho43zkb.pdf.
  • 2. Classification of software quality metrics. (n.d.). Stud.com.ua. Retrieved October 17, 2025, from https://stud.com.ua/128170/informatika/metriki_yakosti_programnogo_zabezpechennya/.
  • 3. The Goal/Question/Metric Paradigm (GQM). (n.d.). Retrieved October 17, 2025, from https://publica.fraunhofer.de/bitstreams/3c4dda1a-1c71-4a86-8b29-9a164bb31037/ download.
  • 4. Lindstrom. (n.d.). The GQM approach is not static but might need refinement. Retrieved Oc­tober 17, 2025, from https://fileadmin.cs.lth.se/serg/old-serg-dok/docs-masterthesis /62_Lindst­rom_draft.pdf.
  • 5. Requirements management measures help organizations to understand, control and assess requirements management process. (2007). ACIT. Retrieved October 17, 2025, from https://www.acit2k.org/ACIT2007/Proceeding/204.pdf.
  • 6. Stability ratio of requirements. (н.д.). QAGroup. Retrieved October 17, 2025, from https://qagroup.com.ua/publications/what-is-metrics/.
  • 7. Common challenges and pitfalls implementing quality metrics. (n.d.). SafetyCulture. Ret­rie­ved October 17, 2025, from https://safetyculture.com/topics/quality-metrics/.
  • 8. Role of LLMs and AI in software requirements quality metrics future research. (2025). Fron­tiers in Computer Science. https://doi.org/10.3389/fcomp.2025.1519437.
  • 9. Challenges in quantifying natural language requirements understanding and its impact on metrics. (n.d.). Shelf.io. Retrieved October 17, 2025, from https://shelf.io/blog/challenges-and-considerations-in-nlp/.
  • 10. Major challenges of Natural Language Processing. (n.d.). GeeksforGeeks. Retrieved October 17, 2025, from https://www.geeksforgeeks.org/nlp/major-challenges-of-natural-language-processing/.
  • 11. Case study compilation - How software testing companies maximize client ROI. (n.d.). MoldStud. Retrieved October 17, 2025, from https://moldstud.com/articles/p-case-study-compilation-how-software-testing-companies-maximize-client-roi.
  • 12. Formal specification in software engineering. (n.d.). Medium. Retrieved October 17, 2025, from https://afzalbadshah.medium.com/formal-specification-in-software-engi­nee­ring-b66852d96fc6.
  • 13. Empirical research on requirements quality: a systematic mapping study. (2022). ResearchGate. https://doi.org/10.1145/1235.
  • 14. An empirical study of software requirements verification and validation techniques along their mitigation strategies. (2015). ResearchGate. Retrieved October 17, 2025, from https://www.researchgate.net/publication/281645652_An_Empirical_Study_of_Software_Requirements_Verification_and_Validation_Techniques_along_their_Mitigation_Strategies.
  • 15. Requirements impact as recorded in the issue reports. (2024). arXiv. https://doi.org/10.48550/arXiv.2402.00594.
  • 16. NLP-assisted software testing_SLR. (n.d.). Queen’s University Belfast. Retrieved October 17, 2025, from https://pureadmin.qub.ac.uk/ws/files/219213040/NLP_assisted_software_tes­ting_ SLR_Mar_01.pdf.
  • 17. Research on applying ML and NLP techniques to the classification of non-functional requirements and performance metrics. (2020). University of Manchester. Retrieved October 17, 2025, from https://research.manchester.ac.uk/files/183513868/2020_12_16_NLP4RE_ Ac­cepted.pdf.
  • 18. Traceability, the ability to trace relevant software artifacts to support reasoning about the quality of the software. (2024). arXiv. https://doi.org/10.48550/arXiv.2405.10845.
  • 19. Hileta I. V. Kiselyk R.O., Karnauchov V.A, Komynar T.N. Overview of modern approaches to software quality assessment Scientific Notes No. 1 (70) / 2025. P. 72--82, doi: 10.32403/1998-6912-2025-1-70-72-82.