<p>Efficient task scheduling is a critical aspect of cloud computing, aiming to optimize resource utilization and minimize processing delays. However, conventional metaheuristic algorithms often suffer from premature convergence, leading to suboptimal scheduling outcomes. To address this challenge, this study proposes a novel multi-objective task scheduling framework designed for big data applications in cloud environments. The framework utilizes an Enhanced Golden Jackal Optimization (EGJO) algorithm, integrated with Opposition-Based Learning (OBL), to enhance population diversity and accelerate convergence. The proposed approach optimizes a multi-objective function encompassing key factors such as cost, makespan, energy efficiency, and resource utilization. Experimental evaluations conducted using diverse benchmarks and performance metrics demonstrate that the proposed framework significantly outperforms existing methods in terms of scheduling efficiency, adaptability, and system reliability within heterogeneous and dynamic cloud environments.</p>