Performance Optimization of Self-represented Systems Supporting Run-time Meta-modeling

Abstract

Fully self-representing architecture, which supports parallel evolution of meta-models and models when the system running, was proposed by integrating runtime meta-modeling and run-time modeling approaches in current research. Secondly, we focused on the study of the unified structure method, which was put forward to represent in-memory objects at different levels of abstraction and can significantly simplify caching mechanism. After that, a 3-level caching system based on web application was proposed in current research as well. In the end, the availability of the proposed approach was verified from aspects of performance analysis and practical application.