Analyzing the Performance of Distributed Web Systems Within an Educational Assessment Framework

This paper examines the usage of genetic algorithms for performance analysis and optimization in distributed web systems. The selected system for distributed systems performance analysis pertains to educational assessment domains. Genetic algorithms (GA) offer a promising approach for automating the optimization process. In this study, the GA is used to generate educational assessment tests within an educational framework. The assessment tests are formed of items stored in data structures and used in the generation process. This process is delimited by several requirements of the assessment objectives, such as the degree of difficulty or solving time. A typical GA approach leads to optimised sequences of items or tests, whereas the usage of distribution within the process can surpass usual GA issues, such as local optimisation. The distribution consists in the generation of the assessment tests on several nodes within a network, where each node has generative tasks. Moreover, they have been successfully applied and solved a wide range of problems, including scheduling, routing, and load balancing in distributed systems. This article also presents a comparative analysis of GA performance: centralized GA and distributed GA. Centralized GAs run on a single computer, while distributed GAs run on multiple computers. This work represents an important step in the understanding and application of genetic algorithms in the context of distributed web systems, providing a solid foundation for future research in this area. The conclusions of this paper demonstrate the effectiveness of genetic algorithms in developing the performance of distributed systems.