Applications of program understanding and rule-based quality assurance to Slam II simulation programs.

• Main Author: Wendt, N. Rodney.
• Other Authors: Oren, Tuncer
• Format: Others
• Published: University of Ottawa (Canada) 2009
• Subjects: Computer Science.
• Online Access: http://hdl.handle.net/10393/6893; http://dx.doi.org/10.20381/ruor-11510

With the advance of time, our inventory of simulation programs has and continues to accumulate. To maximize the return on our investment of time and money into these software systems, it is advantageous for us to reuse software components as much as possible. For example, previously engineered simulation models can often be reused and exercised under a new set of experimental conditions. Before a software component can be reused, the analyst must learn and understand its functionality. This learning process is often made unnecessarily difficult due to incomplete documentation. Another contributing factor is the complexity brought about by interacting directly with the program code. Furthermore, when it comes time to make updates to the code, the potential arises for semantic and syntactic errors to work their way into the program. Knowledge-based program understanding systems with built in quality assurance can be used as an environment for simplifying the learning and the update processes, while ensuring an acceptable degree of quality has been maintained during the update process. This thesis discusses program understanding and quality assurance issues related to the Slam II programming language and discusses the architecture of E/Slam (Elucidation of Slam II programs). E/Slam is a knowledge-based program understanding system with built-in quality assurance ability.