The objective of this paper is to implement an "adaptive modeling support system (AMSS)", which includes modeling, model understanding and modeling adaptive improvement. The paper identifies the knowledge in AMSS which is based on visualization cluster analysis (VCA) to support the iterative cycle, and focuses on studying the learning methods to obtain such knowledge and architecture to support the capabilities. OSP (the Oil Stabilization Process) is used to demonstrate the functionality of AMSS that results in an effective process modeling tool with high solution quality and consistency.