A Transformer-Guided Semantic and Swarm-Based Optimization Model for High Performance Database Query Processing using TSPACO

Abstract

Query optimization is a crucial operation within a current database system used to enhance performance of the database in terms of speed, accuracy and efficiency of resource usage in a database system when dealing with large scale relational information. The common limitations of the existing query optimization techniques include high execution time, high cost of computation and also inability to adapt to complex SQL queries. This research aims at coming up with an intelligent and dynamic query optimization framework that has the capability of producing the best execution plans in a manner that is both accurate and low in computation costs. A Transformer based Semantic and Hybrid PSO-ACO (TSPACO) query optimization model is proposed as a result of this aim, integrating semantic query analysis, row and table filtering with a hybrid optimization combining Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO). The semantic analysis step determines the tables and rows that are relevant whereas the hybrid optimization step chooses the optimal execution path by balancing between exploration and exploitation strategies. To test the proposed method, Python is used and tested with relational query data to quantify execution time, accuracy, cost of computations and throughput. The proposed TSPACO method is demonstrated to outperform the existing methods with better performance as indicated in experimental results with a query accuracy of 97% at a shorter time and lower cost of computation. The findings affirm that the proposed framework is an efficient and scalable intelligent query optimization solution to large database systems.