Performance Verification Mechanism for Adaptive Assessment e-Platform and e-Navigation Application

Abstract

Adaptive assessment e-platform is being promoted in the world to make teachers understand students’ e-learning performance on the Internet. However, system's load testing for an adaptive assessment is a very important issue during development of such an e-platform. In this paper, we have adopted the genetic fuzzy markup language (GFML) to infer the performance of an adaptive assessment e-platform. Firstly, we collected the data and information of the e-platform loading in two different mechanisms. With the collected data, the proposed CPU usage calculation mechanism is first implemented to acquire the CPU usage information from the screenshot of Ganglia. Next, we used the fuzzy c-means (FCM) clustering mechanism to construct the knowledge base according to the collected data. Then, number of threads, constant timer, MySQL parameter, CPU usage, and testing time of the e-platform were utilized to infer the e-platform load performance. Finally, the genetic learning algorithm was utilized to learn the knowledge and rule base to optimize the proposed approach. From these experimental results, the proposed method is feasible for verifying the performance of an adaptive assessment e-platform. In the future, the adaptive assessment e-platform can be utilized to e-Navigation systems and applications.