Abstract: The architecture of the Smart Test is described. Instead of writing a script, the Tester designs a functional model of the system to be tested, such as an application. For example, the Tester would model the functions of the system to be tested, such as a main window, a menu bar, drop-down menus, specialized windows etc. The more complete the model, the better the chance testing will cover existing function. The Tester would also define any facts that the model might need (for example, the name of the file to be opened and saved). The Tester then defines any goals or subgoals to be attained. A goal might be saving the file. Then the rules under which the model will operate are defined. One rule might be if the file (named as a fact) has its date and/or time changed, then the goal of saving the file was reached and the test will end.

Abstract: An automated software test is provided which includes a functional model of a system to be tested. The automated software test is utilized to operate a system under test in accordance with specified facts, goals and rules. Quasi-random actions are taken within the system in accordance with specified rules and facts until a defined goal has been accomplished. Training the automated software test is accomplished by specifying a particular goal, i.e. identifying a particularly known defect, and thereafter running the test in a quasi-random fashion until the particular goal has been achieved. The number and nature of actions required to achieve that goal are logged and the process is then repeated until the shortest path required to achieve that goal has been determined. The log of actions which eventually reach a particularly defect may also be utilized a probable cause tree structure for future analysis.

Abstract: A preprocessor translates problem descriptions in text format to a numeric form, usable by a neural network. Problem descriptions and associated solutions are used as training sets for the neural network. From a word domain collected from a number of problem descriptions, each character used in the problem domain is assigned a character frequency value. These frequency values are used to translate each problem description to an input vector, where each input value represents a character and each input value is associated with an input node of the neural network. In addition to being associated with a frequency value, each character is scaled and normalized to improve accuracy of the neural network's recall capabilities.