Gaining Insight from Crash Reports and Bug Reports

This project seeks to address the flood of bugreports and crashreports that software companies must address when they build consumerfacing software. Organizations can easily face 50,000 crash reports per day. Developers cannot manually sift through them all. The goal of this work is to exploit modern search engine technology and crash report representation to enable the fast and efficient querying, clustering, and reporting of crash reports so that developers do not have to group crash report manually. This enables an organization, such as our partner organization, to be aware of new issues arising as well as to efficiently address
issues that endusers are facing during deployment. The same is true of bug reports. Multiproduct bug reports are numerous and hard to address manually. Partner organizations benefit from increased software quality assurance efficiency, Canadian industry benefits from the release of powerful opensource tools, and interns benefit from industrially relevant training.

Faculty Supervisor:

Abram Hindle

Student:

Joshua Charles Campbell

Partner:

BioWare

Discipline:

Computer science

Sector:

Digital media

University:

University of Alberta