Specifying operational profiles for modules

Abstract

We describe a technique for tional profiles for modules. specifying operaThe technique is more general than those of the current literature and allows more accurate specification of nlodu]e usage. We also outline an algorithm for allsomatically generating random test cases from any such operational profile specification for a module, such that the test cases correspond to a random sampling of the module's input in actual operation. Operational-based statistical estimations, such as operational reliability, may be more meaningful when our specification method and generation algorithm are used, because our method permits more precise specifications than do other methods in the current literature.

Using a pilot study to derive a GUI model for automated testing

“…Numerous researchers have developed techniques that employ state machine models [Clarke 1998;Chow 1978;Esmelioglu and Apfelbaum 1997;Bernhard 1994;Shehady and Siewiorek 1997;White and Almezen 2000], grammars [Imanian 2005;von Mayrhauser and Crawford-Hines 1993;von Mayrhauser et al 1994;Maurer 1990;Auguston et al 2005], AI planning Howe et al 1997;Scheetz et al 1999;Leow et al 2004], genetic algorithms [Kasik and George 1996], probabilistic models [Woit 1998;1993;Whittaker 1992;Whittaker and Thomason 1994], and graph-traversal techniques ] to generate various types of sequences. This section presents some of these techniques (emphasizing on their application to GUIs whenever possible), a discussion of GUI testing tools, and the relationship between our work and unit testing of object-oriented programs [Jorgensen and Erickson 1994;, and delta debugging [Zeller 2002].…”

“…Probabilistic Models: Several probabilistic models have been used to encode software usage information (GUI usage may be encoded as sequences of events) [Woit 1998;1993;Whittaker 1992;Whittaker and Thomason 1994]. These models are used to generate test cases that will exercise the software in a fashion typical of actual operation.…”

“…Whittaker et al [Whittaker and Thomason 1994;Whittaker 1992] describes a method for statistical testing based on a Markov chain model of software usage, which allows test input sequences to be generated from multiple probability distributions. Woit [Woit 1993;] extends this idea to a conditionalevent usage testing (CEUT), which retains the benefits of traditional operational profile testing but is not limited in the class of systems to which it applies. In CEUT, the expected usage of the software is modeled in such a way as to allow specification of conditions upon expected input event sequences.…”

Automated gui testing guided by usage profiles

“…Our previous work investigated replaying usage profiles to perform regression testing on applications, but did not employ usage profiles to create a model of the application and test the application based on that model [12]. Many of the previous approaches to testing with usage profiles rely on manually generating profiles based on a priori knowledge of the application [23,24]. The overall goal of testing using these methods is to develop and implement fully automated methods in which user interactions are captured, the application or application model is generated automatically from the binary code, and test cases are developed and executed.…”

“…Usage profiles (called operational profiles [24] or session data [6] in some literature) can be obtained by capturing event sequences users execute on the GUI application. Because they are gathered from actual usage of the software, any faults detected by using them represent faults likely to be encountered by users in the field.…”