Techniques for Reducing the Complexity of Object-Oriented Execution Traces.
Abdelwahab Hamou‐Lhadj, Timothy C. Lethbridge · 2003
Understanding the behavior of object-oriented systems is almost impossible by merely performing static analysis of the source code. Dynamic analysis approaches are better suited for this purpose. Run time information is typically represented in the form of execution traces that contain object interactions. However, traces can be very large and hard to comprehend. Visualization tools need to implement efficient filtering techniques to remove unnecessary data and present only information that adds value to the comprehension process. This paper addresses this issue by presenting different filtering techniques. These techniques are based on removing utility methods and the use of object-oriented concepts such as polymorphism and inheritance to hide low-level implementation details. We also experiment with 12 execution traces of an object-oriented system called WEKA and study the gain attained by these filtering techniques