February 23, 2018  - CINI ITEM National Lab - Centri Comuni  Complesso Universitario Montesantangelo

h: 10.00 - 12.00


Mozilla, London, UK

Improving software quality through machine learning and data mining

How we are using machine learning and data mining techniques at Mozilla to improve software quality and engineering practices: automatic understanding of groups of crash reports with contrast set learning; improving clustering of crash reports using Natural Language Processing techniques; empirical studies of software processes at Mozilla (analyzing the processes around uplifts / backports, the review processes, how bugs caused by third-party software affect the products and how they are managed); automatically detecting web-compatibility issues (with virtually infinite test oracles) using Convolutional Neural Networks; studying and improving processes around flaky tests management (e.g. using code coverage information). Future directions.

Share This

S5 Box

Cini Single Sign ON

This site only stores technical/functional cookies. If you want to know more, go to the Cookie Policy section.