Title: AI-Driven Adaptive Sorting Techniques for Large-Scale Data Processing
Authors: Ibrahim Iyad Ibrahim Abd Al Jawwad, Samy S. Abu-Naser
Volume: 10
Issue: 7
Pages: 52-57
Publication Date: 2026/07/28
Abstract:
Efficient sorting is a fundamental requirement in large-scale data processing, directly influencing the performance of distributed computing frameworks, database management systems, and cloud-based analytics platforms. Although traditional sorting algorithms such as Quicksort, Merge Sort, and Heap Sort provide well-established theoretical guarantees, their performance varies considerably depending on the statistical characteristics of input data. Consequently, selecting a single sorting algorithm for all datasets often results in suboptimal computational efficiency. This paper proposes an Artificial Intelligence (AI)-driven adaptive sorting framework that employs machine learning techniques to analyze dataset characteristics and dynamically select the most appropriate sorting algorithm before execution. The proposed framework extracts lightweight statistical features, including data entropy, duplication ratio, degree of sortedness, and value distribution, and uses a Random Forest classifier to predict the optimal sorting strategy. Experimental simulations conducted on synthetic and real-world datasets demonstrate that the proposed framework improves execution time by 15–20% for datasets exceeding 10710^7107 records while maintaining minimal prediction overhead. Furthermore, the adaptive approach reduces CPU utilization and improves overall throughput in distributed computing environments. These findings suggest that integrating AI with algorithm selection offers a promising direction for developing intelligent, self-optimizing data processing systems capable of adapting to heterogeneous workloads.Keywords: Artificial Intelligence, Adaptive Sorting, Big Data, Machine Learning, Computational Complexity, Data Processing.