Prof. Dr. Jeff Offutt
Dr. Jeff Offutt is a Professor of Software Engineering at George Mason University. He has published over 175 refereed research papers (h-index of 59), and invented numerous widely used test techniques. He leads the SPARC educational project and the widely used muJava project. Offutt is editor-in-chief of Wiley’s journal of Software Testing, Verification and Reliability, helped found the IEEE International Conference on Software Testing, Verification, and Validation (ICST), and is co-author of Introduction to Software Testing. He was awarded the George Mason University Teaching Excellence Award, Teaching With Technology, in 2013, was a Mason Outstanding Faculty member in 2008 and 2009, and his IEEE Software paper in software engineering education was chosen by ACM as a notable paper for 2014. His current projects include, software engineering education, automatic repair of SQL queries, critical systems testing, testing and analysis of web and mobile applications, and model-based testing. Offutt received the PhD in Computer Science from the Georgia Institute of Technology in 1988. He is on the web at https://cs.gmu.edu/~offutt/
From Spec-based testing to Test Automation and Beyond
Abstract: Specification-based testing has roots in the mid-1970s, when Huang suggested testing telecommunications software by covering nodes in finite-state machines. As formal languages were developed in the 1980s and 1990s, the field invented more techniques to generate tests from those formalisms. However, many practicing engineers were not comfortable with formal languages, thus scientists developed informal modeling languages. These innovations led to model-based testing (MBT). MBT generates abstract tests in terms of a model, then translates the abstract tests to concrete tests that can be run on an implementation. Engineers use test criteria to choose test suites that are effective and of a reasonable size. They also automate concrete tests so they can run efficiently and repeatedly.
This talk will first summarize this evolution, and then discuss the ongoing disagreement between generating tests from formal languages and from informal languages. Models can be created directly from modeling languages, or from any software artifact. Although some view MBT as only applying when tests are derived from models that were explicitly created to describe software behavior, I suggest this narrow view represents echos of past steps in the evolution of MBT. Models for testing can be generated from any software artifact, including natural language requirements, comments, and source code. The latter part of this talk will present some of my recent research into automating tests within MBT, including the mapping problem and the oracle problem. Finally, I will talk about some current technical problems I am working on.
Dr. Brice Morin
Dr. Brice Morin is a Senior Research Scientists at SINTEF Digital. He has published over 60 refereed papers (h-index of 21). His research focuses on providing sound and effective Model-Driven techniques for the design, implementation and operation of software-intensive systems in general, with a particular emphasis on Microservices and Internet-of-Things (IoT) systems. He is currently working on the H2020 STAMP project, a project focused on Software Testing AMPlification for microservices, where he and his colleagues have developed open-source solutions for configuration amplification of Docker-based systems. He is on the web at http://brice-morin.info/
Configuration testing of Docker-based Microservices
Abstract: To be added.