Synopsis of Possible Collaboration in the Area of Software Debugging

Software debugging has been regarded as one of the challenges in the field of software engineering. In practice, traditional debugging needs extensive labor and time in order to locate bugs in a typical software program. A typical programmer needs to set breaking points and execute the buggy program on several inputs. At the same time, he needs to examine the program states and try to locate the cause that leads to program failure. Automated software fault localization has been recognized as one of the approaches which could improve efficiency in locating software bugs. Dynamic analysis [1], [2], [3] is one of the debugging schemes where software code has been executed with test cases (test inputs) to reveal potential software faults. One of the approaches of dynamic analysis is known as program spectra. Program information such as the statements that are executed in each test case can be gathered. These information can be used to rank the parts of the program according to how likely it is that they contain a bug. Ranking is done by sorting based on the value of a numeric function or better known as ranking metric applied to the data for each part of the program.