Managerial decision analysis using small language models

Authors

  • Ksenia V. Andrenko Brest State Technical University, 267 Maskowskaja Street, Brest 224017, Belarus
  • Aliaksandr A. Kroshchanka Brest State Technical University, 267 Maskowskaja Street, Brest 224017, Belarus
  • Vladimir A. Golovko Brest State Technical University, 267 Maskowskaja Street, Brest 224017, Belarus
  • Yury B. Krapivin Brest State Technical University, 267 Maskowskaja Street, Brest 224017, Belarus

Keywords:

small language model, DistilBERT, management decision analysis, neural network, knowledge distillation

Abstract

This paper proposes a method for the automatic assessment of managerial decision quality based on free-form textual responses to Russian-language business cases using a fine-tuned DistilBERT model. The novelty of the study lies in a context-dependent classification algorithm with concatenated input (case description + response) and a post-processing technique that explicitly links predictions to specific managerial competencies or red flags. For the first time, high-accuracy evaluation of managerial decisions in Russian (accuracy = 0.986 8, F1-score = 0.987 0) with interpretable competency-based feedback has been achieved, unlike prior work focused on general text or sentiment analysis. The approach enables the development of automated systems for managerial assessment, personnel selection and personalised leadership development in human resources and corporate training.

Author Biographies

  • Ksenia V. Andrenko, Brest State Technical University, 267 Maskowskaja Street, Brest 224017, Belarus

    trainee lecturer at the department of intelligent information technologies, faculty of electronic and information systems

  • Aliaksandr A. Kroshchanka, Brest State Technical University, 267 Maskowskaja Street, Brest 224017, Belarus

    PhD (engineering); associate professor at the department of intelligent information technologies, faculty of electronic and information systems

  • Vladimir A. Golovko, Brest State Technical University, 267 Maskowskaja Street, Brest 224017, Belarus

    doctor of science (engineering), full professor; head of the department of intelligent information technologies, faculty of electronic and information systems

  • Yury B. Krapivin, Brest State Technical University, 267 Maskowskaja Street, Brest 224017, Belarus

    PhD (engineering); associate professor at the department of intelligent information technologies, faculty of electronic and information systems

References

  1. Hinton G, Vinyals O, Dean J. Distilling the knowledge in a neural network. arXiv:1503.02531 [Preprint]. 2015 [cited 2024 December 19]: [9 p.]. Available from: https://arxiv.org/abs/1503.02531.
  2. Prema V, Elavazhahan V. Sculpting DistilBERT: enhancing efficiency in resource-constrained scenarios. In: Dwivedi RK, Saxena AK, Sharma R, Bhardwaj S, Khattri V, editors. Proceedings of the 2023 12th International conference on system modeling & advancement in research trends (SMART); 2023 December 22–23; Moradabad, India. [S. l.]: Institute of Electrical and Electronics Engineers; 2023. p. 251–256. DOI: 10.1109/SMART59791.2023.10428568.
  3. Mukherjee A, Umme Salma M. Resume ranking and shortlisting with DistilBERT and XLM. In: Institute of Electrical and Electronics Engineers. Proceedings of the 2024 IEEE International conference for women in innovation, technology and entrepreneurship (ICWITE); 2024 February 16–17; Bangalore, India. [S. l.]: Institute of Electrical and Electronics Engineers; 2024. p. 301–304. DOI: 10.1109/ICWITE59797.2024.10502523.
  4. Jingga K, Hidayaturrahman H. The utilization of DistilBERT features space for topics suggestion in the context of learning content. In: Jusman Y, Mat Isa NA, Sukamta, Ejaz W, Ismail AH, Bhattacharjya A, et al., editors. Proceedings of the 2023 International workshop on artificial intelligence and image processing (IWAIIP); 2023 December 1–2; Yogyakarta, Indonesia. [S. l.]: Institute of Electrical and Electronics Engineers; 2023. p. 243–247. DOI: 10.1109/IWAIIP58158.2023.10462787.
  5. Islam MT, Parvin F, Sazan SA, Amir TB. Comparative analysis of sentiment classification on IMDB 50k movie reviews: a study using CNN, LSTM, CNN-LSTM, and BERT models. In: Institute of Electrical and Electronics Engineers. Proceedings of the 2024 IEEE International conference on power, electrical, electronics and industrial applications (PEEIACON); 2024 September 12–13; Rajshahi, Bangladesh. [S. l.]: Institute of Electrical and Electronics Engineers; 2024. p. 512–517. DOI: 10.1109/PEEIACON63629.2024.10800035.
  6. Otani N, Bhutani N, Hruschka E. Natural language processing for human resources: a survey. In: Chen W, Yang Y, Kachuee M, Fu X-Y, editors. Proceedings of the 2025 Conference of the nations of the Americas chapter of the Association for Computational Linguistics: human language technologies; April 2025; Albuquerque, USA. Volume 3, Industry track. Kerrville: Association for Computational Linguistics; 2025. p. 583–597. DOI: 10.18653/v1/2025.naacl-industry.47.
  7. Kang Y, Cai Z, Tan C-W, Huang Q, Liu H. Natural language processing (NLP) in management research: a literature review. Journal of Management Analytics. 2020;7(2):139–172. DOI: 10.1080/23270012.2020.1756939.
  8. Schmiedel T, Müller O, vom Brocke J. Topic modeling as a strategy of inquiry in organizational research: a tutorial with an application example on organizational culture. Organizational Research Methods. 2019;22(4):941–968. DOI: 10.1177/1094428118773858.
  9. Jia J, Liang W, Liang Y. A review of hybrid and ensemble in deep learning for natural language processing. arXiv:2312.05589v2 [Preprint]. 2024 [cited 2024 December 19]: [23 p.]. Available from: https://arxiv.org/abs/2312.05589v2.
  10. Davenport T, Guha A, Grewal D, Bressgott T. How artificial intelligence will change the future of marketing. Journal of the Academy of Marketing Science. 2020;48(1):24–42. DOI: 10.1007/s11747-019-00696-0.
  11. Golestani A, Masli M, Shami NS, Jones J, Menon A, Mondal J. Real-time prediction of employee engagement using social media and text mining. In: Wani MA, Kantardzic M, Sayed-Mouchaweh M, Gama J, Lughofer E, editors. Proceedings of the 17th IEEE International conference on machine learning and applications (ICMLA); 2018 December 17–20; Orlando, USA. [S. l.]: Institute of Electrical and Electronics Engineers; 2018. p. 1383–1387. DOI: 10.1109/ICMLA.2018.00225.
  12. Jamuna KM. Leveraging natural language processing to analyze employee feedback for enhanced HR insights. International Journal of Scientific Research in Engineering and Management. 2024;8(11):1–6. DOI: 10.55041/IJSREM39115.
  13. Leidner JL, Stevenson M. Challenges and opportunities of NLP for HR applications: a discussion paper. arXiv:2405.07766 [Preprint]. 2024 [cited 2024 December 19]: [10 p.]. Available from: https://arxiv.org/abs/2405.07766.
  14. Zhang M, Jensen KN, Sonniks SD, Plank B. SkillSpan: hard and soft skill extraction from English job postings. In: Proceedings of the 2022 Conference of the North American chapter of the Association for Computational Linguistics: human language technologies; 2022 July 10–15; Seattle, USA. Stroudsburg: Association for Computational Linguistics; 2022. p. 4962–4984. DOI: 10.18653/v1/2022.naacl-main.366.
  15. Uma VR, Velchamy I, Upadhyay D. Recruitment analytics: hiring in the era of artificial intelligence. In: Tyagi P, Chilamkurti N, Grima S, Sood K, Balusamy B, editors. The adoption and effect of artificial intelligence on human resources management. Part A. Bingley: Emerald Publishing Limited; 2023. p. 155–174 (Özen E, Grima S, editors. Emerald studies in finance, insurance, and risk management; volume 7).
  16. França TJF, Mamede HS, Barroso JMP, dos Santos VMPD. Artificial intelligence applied to potential assessment and talent identification in an organisational context. Heliyon. 2023;9(4):e14694. DOI: 10.1016/j.heliyon.2023.e14694.
  17. Devlin J, Chang M-W, Lee K, Toutanova K. BERT: pre-training of deep bidirectional transformers for language understanding. In: Burstein J, Doran C, Solorio T, Kumar R, Loukina A, Morales M, et al., editors. Proceedings of the 2019 Conference of the North American chapter of the Association for Computational Linguistics: human language technologies; 2019 June 2–7; Minneapolis, USA. Volume 1, Long and short papers. Stroudsburg: Association for Computational Linguistics; 2019. p. 4171–4186. DOI: 10.18653/v1/N19-1423.
  18. Gardazi NM, Daud A, Malik MK, Bukhari A, Alsahfi T, Alshemaimri B. BERT applications in natural language processing: a review. Artificial Intelligence Review. 2025;58(6):166. DOI: 10.1007/s10462-025-11162-5.
  19. Davoodi L, Mezei J, Heikkilä M. Aspect-based sentiment classification of user reviews to understand customer satisfaction of e-commerce platforms. Electronic Commerce Research. 2026;26(2):1417–1459. DOI: 10.1007/s10660-025-09948-4.
  20. Abdel-Salam S, Rafea A. Performance study on extractive text summarization using BERT models. Information. 2022;13(2):67. DOI: 10.3390/info13020067.
  21. Lan Z, Chen M, Goodman S, Gimpel K, Sharma P, Soricut R. ALBERT: a lite BERT for self-supervised learning of language representations. arXiv:1909.11942v6 [Preprint]. 2020 [cited 2024 December 19]: [17 p.]. Available from: https://arxiv.org/abs/1909.11942v6.
  22. Barbon RS, Akabane AT. Towards transfer learning techniques – BERT, DistilBERT, BERTimbau, and DistilBERTimbau for automatic text classification from different languages: a case study. Sensors. 2022;22(21):8184. DOI: 10.3390/s22218184.
  23. Sanh V, Debut L, Chaumond J, Wolf T. DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter. arXiv:1910.01108v4 [Preprint]. 2020 [cited 2024 December 19]: [5 p.]. Available from: https://arxiv.org/abs/1910.01108.
  24. Andrenko KV, Kroshchanka AA, Golovko OV. The use of distilled large language models to determine the sentiment of a text. In: Golenkov VV, Azarov IS, Golovko VA, Gordey AN, Guliakina NA, Krasnoproshin VV, et al., editors. Open semantic technologies for intelligent systems. Issue 9. Minsk: Belarusian State University of Informatics and Radioelectronics; 2025. p. 229–234.
  25. Liu Y, Ott M, Goyal N, Du J, Joshi M, Chen D, et al. RoBERTa: a robustly optimized BERT pretraining approach. arXiv:1907.11692 [Preprint]. 2019 [cited 2024 December 19]: [13 p.]. Available from: https://arxiv.org/abs/1907.11692.
  26. Jiao X, Yin Y, Shang L, Jiang X, Chen X, Li L, et al. TinyBERT: distilling BERT for natural language understanding. In: Cohn T, He Y, Liu Y, editors. Findings of the Association for Computational Linguistics: EMNLP-2020; 2020 November 16–20. Stroudsburg: Association for Computational Linguistics; 2020. p. 4163–4174. DOI: 10.18653/v1/2020.findings-emnlp.372.
  27. Golovko V, Kroshchanka A, Treadwell D. The nature of unsupervised learning in deep neural networks: a new understanding and novel approach. Optical Memory and Neural Networks. 2016;25(3):127–141. DOI: 10.3103/S1060992X16030073.
  28. Golovko V, Kroshchanka A, Turchenko V, Jankowski S, Treadwell D. A new technique for restricted Boltzmann machine learning. In: Institute of Electrical and Electronics Engineers. Proceedings of the 2015 IEEE 8th International conference on intelligent data acquisition and advanced computing systems: technology and applications (IDAACS); 2015 September 24–26; Warsaw, Poland. Volume 1. [S. l.]: Institute of Electrical and Electronics Engineers; 2015. p. 182–186. DOI: 10.1109/IDAACS.2015.7340725.
  29. Andrenko KV, Kroshchanka AA, Golovko VA, Krapivin YuB. The use of neural networks for decision-making analysis. In: Ablameyko SV, Kazachonak VU, Kurbackij AN, Krasnoproshin VV, editors. Information systems and technologies. Proceedings of the 11th International scientific congress on computer science (CSIST-2025); 2025 October 29–31; Minsk, Belarus. Part 2. Minsk: Belarusian State University; 2025. p. 118–125. Russian.

Downloads

Additional Files

Published

2026-05-21

Issue

No. 1 (2026): Journal of the Belarusian State University. Mathematics and Informatics

Section

Data Science

License

Copyright (c) 2026 Journal of the Belarusian State University. Mathematics and Informatics

Creative Commons License

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

The authors who are published in this journal agree to the following:

  1. The authors retain copyright on the work and provide the journal with the right of first publication of the work on condition of license Creative Commons Attribution-NonCommercial. 4.0 International (CC BY-NC 4.0).
  2. The authors retain the right to enter into certain contractual agreements relating to the non-exclusive distribution of the published version of the work (e.g. post it on the institutional repository, publication in the book), with the reference to its original publication in this journal.
  3. The authors have the right to post their work on the Internet (e.g. on the institutional store or personal website) prior to and during the review process, conducted by the journal, as this may lead to a productive discussion and a large number of references to this work. (See The Effect of Open Access.)

How to Cite

[1]
Andrenko, K.V. et al. 2026. Managerial decision analysis using small language models. Journal of the Belarusian State University. Mathematics and Informatics. 1 (May 2026), 121–129.