An Engineering Approach to Explainable Artificial Intelligence for SQL Query Optimization: Design and Evaluation of the OptiMind Framework
Al-Noor Journal of Engineering Management and Computer Science · 14.06.2026 · 87-105
DOI: 10.71229/fxy9xp82 ↗Kurz gefasst
Fachartikel in der Zeitschrift Al-Noor Journal of Engineering Management and Computer Science, online erschienen am 14.06.2026. Verfasst von Hanan. Die Arbeit nennt 19 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-SA 4.0).
Abstract (Originalsprache)
Query optimization remains a critical challenge in Relational Database Management Systems (RDBMSs) because it directly affects execution efficiency, resource utilization, and overall database performance. Traditional cost-based and rule-based optimizers rely on statistical estimation and heuristic plan enumeration methods that often perform poorly when handling correlated predicates, skewed data distributions, and complex multi-join workloads. Although recent learned query optimization approaches have improved cost estimation accuracy, many existing solutions operate as black-box models with limited interpretability, weak workload adaptability, and insufficient support for database administrator (DBA) analysis and decision-making. This paper presents OptiMind, an explainable intelligent framework for SQL query optimization that integrates semantic SQL analysis, workload-aware behavior profiling, adaptive feature intelligence, and interpretable optimization scoring into a unified recommendation system. Unlike traditional optimization pipelines, OptiMind emphasizes transparency by generating human-readable optimization recommendations through clause dependency analysis, operator behavior profiling, workload-aware feature ranking, and feature-attribution scoring. The framework improves both optimization accuracy and explainability while maintaining adaptability across varying workload environments and query structures. OptiMind was evaluated using benchmark and enterprise workloads, including JOB, IMDB, SQLShare, Spider, and enterprise traces. Experimental results demonstrate that the framework reduces mean absolute error (MAE) by 18.4%–31.7% compared with PostgreSQL’s native optimizer and achieves execution speedups ranging from 1.31× to 2.14× on complex multi-join queries. Additionally, OptiMind achieved an interpretability score of 4.51/5.0, confirming strong explanation quality, recommendation usefulness, and practical effectiveness for transparent, workload-adaptive, and explainable SQL query optimization in modern database systems.
Zitieren
APA
Hanan (2026). An Engineering Approach to Explainable Artificial Intelligence for SQL Query Optimization: Design and Evaluation of the OptiMind Framework. Al-Noor Journal of Engineering Management and Computer Science, 87-105. https://doi.org/10.71229/fxy9xp82
Vancouver
Hanan. An Engineering Approach to Explainable Artificial Intelligence for SQL Query Optimization: Design and Evaluation of the OptiMind Framework. Al-Noor Journal of Engineering Management and Computer Science. 2026:87-105. doi:10.71229/fxy9xp82
BibTeX
@article{Hanan2026,
author = {Hanan},
title = {An Engineering Approach to Explainable Artificial Intelligence for SQL Query Optimization: Design and Evaluation of the OptiMind Framework},
journal = {Al-Noor Journal of Engineering Management and Computer Science},
year = {2026},
pages = {87--105},
doi = {10.71229/fxy9xp82}
}
EUVIMED