• home
• current students
• graduate
• events

Presenter: Professor Stan Matwin School of Information Technology and Engineering University of Ottawa
Supervisor: Professor M. Miller - Dean of Engineering

Date: Tue, May 13, 2003
Time: 13:30:00 - 14:30:00
Place: Centre Innovative Teaching Building (CIT), Room #110

ABSTRACT

ABSTRACT:

In this presentation we will argue that data mining methods are particularly suitable to support a number of software engineering activities. We will describe our experience with an industrially-based research project in this area. Working with the data from Mitel Corp., we have developed a tool assisting maintenance personnel. For a given module, modified in the maintenance task, the tool suggests other modules, potentially relevant for the task. Our experience shows that, since software maintenance is a knowledge-based activity, then the requisite knowledge may, in many cases, be acquired and generalized from previous experience using machine learning/data mining methods. We show that our maintenance application presents a number of challenges for current data mining methods, and we discuss our solutions to these challenges. We observe that text-based attributes collected from comments and problem reports perform particularly well in predicting the relevance relationship.

Stan Matwin is a professor at the School of Information Technology and Engineering, University of Ottawa. His research is in machine learning, data mining, and their applications. Former president of the Canadian Society for the Computational Studies of Intelligence (CSCSI) and of the IFIP Working Group 12.2 (Machine Learning). Author and co-author of some 120 refereed papers in journals and conferences. Currently, Director of the Information Technology Cluster of the Ontario Research Centre for Electronic Commerce. Member of the Editorial Boards of the Machine Learning Journal and the Intelligent Data Analysis Journal.

Note: Professor Stan Matwin is a candidate for Chair of the Department of Computer Science.