Hybrid Ensemble-Based Effort Prediction for IT Tasks Using Historical Data from the 1C:Enterprise Platform

This paper proposes a hybrid method for predicting the effort required for IT tasks related to the development and maintenance of solutions based on the 1C:Enterprise platform. The method combines expert estimation, structured task attributes, search for semantically similar references using historical data, and an ensemble of machine learning models. A meta-model is used to generate the final prediction by integrating the outputs of the base algorithms with additional features derived from historical data. The experimental verification was carried out on a sample that included more than 12,000 completed tasks. The proposed method demonstrated higher prediction accuracy than individual machine learning models and confirm the effectiveness of combining expert knowledge, semantic retrieval, and historical project data.

1. Мельников А.В. Методика оценки трудоемкости разработки на платформе «1С:Предприятие 8» / А.В. Мельников, Ф.А. Музалевский, Р.А. Солодуха // Судебная экспертиза. – 2026. – № 1 (85). – С. 124–137.

2. Ковалев А.Д. Методы автоматизации сопровождения программного обеспечения на основе интеллектуальной обработки запросов заказчика: автореф. дис. ... канд. техн. наук: 2.3.5 / Ковалев Артем Дмитриевич; Санкт-Петербургский политехнический университет Петра Великого. – Санкт-Петербург, 2023. – 18 с.

3. Fu M. GPT2SP: A Transformer-Based Agile Story Point Estimation Approach / M. Fu, Ch. Tantithamthavorn // IEEE Transactions on Software Engineering. – 2023. – Vol. 49, No. 2. – P. 611–625.

4. Fine-SE: Integrating Semantic Features and Expert Features for Software Effort Estimation / Y. Li, Zh. Ren, Zh. Wang [et al.] // Proceedings of the 46th IEEE/ACM International Conference on Software Engineering (ICSE 2024). – New York: ACM, 2024. – URL: https://ieeexplore.ieee.org/document/10548937 (дата обращения: 18.07.2026).

5. Analysis of Software Effort Estimation by Machine Learning Techniques / M.U. Nisa, M. Saqlain, M. Abid [et al.] // Ingénierie des Systèmes d'Information. – 2023. – Vol. 28, No. 6. – P. 1445–1457.