Department of Computer, Control and Management Engineering

Software testing is an effective practice used to check whether the software products that are being developed match their expected requirements. Modern industrial systems spend a large amount of resources on software testing by running thousands or even millions of tests every day. Many techniques have been presented in the academic literature to tame the costs of software testing. For example, they reorder the test suite or select a subset of it to find software failures quicker. However, most existing techniques are too expensive for handling modern massive systems and depend on artifacts, such as code coverage metrics, that are not commonly available at a large scale. In this talk, we show how information retrieval and data mining techniques can help while applied to software testing problems. In particular, we present two approaches that have comparable effectiveness to state-of-the-art techniques, while providing huge gains in terms of efficiency, as shown by experimental results.

 

Emilio Cruciani got his Ph.D. in Computer Science in 2019 at the Gran Sasso Science Institute. He is currently a postdoctoral researcher in the Big Data Algorithms Group of the Paris-Lodron University of Salzburg. Previously he was a postdoctoral researcher in the Efficient Algorithms Group (University of Salzburg) and the COATI team in INRIA Sophia Antipolis. His research interests include the analysis of dynamics on networks and the design and implementation of scalable algorithms and heuristics for large datasets. His line of research on scalable approaches for software testing is funded by two Facebook Research Awards (2019, 2021).

 

Il will be possible to attend remotely at the following Zoom link: https://uniroma1.zoom.us/j/87006718821?pwd=SG9kemlydHY1UzEvT1JoUnYyQUdiUT09

 

Login to Zoom with uniroma1.it credentials is required before connecting to the link above