Logo Universität Bremen


Universität Bremen - Fachbereich 3 - Informatik


Startseite Detail

A Generic Method for Statistical Testing

Datum: 08.02.2006
Uhrzeit: 17:15-19:00

Ort: MZH 5210

Vortragende(r): Prof. Dr. Marie-Claude Gaudel Homepage - Vortragende(r) (Université de Paris-Sud et Cnrs)

In the area of software testing, numerous methods have been proposed and used for selecting finite test sets and automating this selection. Among these methods some are deterministic and some are randomized (often called “statistical”). In the recent years, the general problem of studying and simulating random processes has particularly benefited from progresses in the area of random generation of
combinatorial structures. In this talk, we explore the idea of using such concepts and tools for statistical
software testing. We describe a generic method for using these tools as soon as there is a graphical
description of the behaviour of the system under test. Uniform generation is used either for drawing paths from the set of execution paths or traces of the system under test, or, more efficiently, among some trace subsets satisfying some coverage conditions. The talk presents the general method and concludes with some experimental results on applying it to C programs and model-based testing.

Marie-Claude Gaudel is a professor at the University of Paris-Sud at Orsay. Before joining UPS, she was a researcher at INRIA, and then in charge of the Software Engineering group at the industrial research center of Alcatel-Alsthom (Marcoussis, France). Her research interests are in the areas of software: formal methods, program robustness, testing and certification. Her main achievements are:
• Development of formal specification languages and environments motivated and validated by industrial
case studies (telephone, train control, and nuclear plant…), or reference case studies (Unix File
Management System, Production Cell, Steam Boiler Control System…)
• Development of a theory and a tool for tests selection from formal specifications; application to several
formal methods and case studies.
• Automation of statistical software testing; assessment of ultra-reliability

Ansprechpartner(in) / Einladende(r):
Prof. Dr. Bernd Krieg-Brückner