Model Tracing of Coding Styles of Programmers: A Formative Approach

Candidate: Shilpi Rao
Type: Master of Science (MSc), School of Interactive Arts and Technology
Date: August 10, 2007
Senior Supervisor: Dr. Marek Hatala, Associate Professor
Thesis: Download Thesis Document

Abstract

Programming Style refers to the ability to follow code conventions, to engineer code in a disciplined manner, to systematically debug code, to optimize code delivery through appropriate settings in the IDE (Integrated Development Environment), to regulate completion rates and quality of programming tasks, and finally to efficiently collaborate with other programmers and resources. This research investigates whether programming styles of individual programmers can be computationally recognized; If styles can be recognized by the machine, can they then be regulated so that programmers can reflect on their own programming styles; finally, can a mixed-initiative computational mechanism assist programmers to identify good programming styles and repair bad programming habits. My research focuses on a real-time architecture called MICE (Mixed-Initiative Coding Environment) that I have developed to tackle this research question.

Keywords: Ontology, Rules, Human-Computer Interaction, Mixed-Initiative Interactions, Self-Regulated Learning, Model-Tracing, Intelligent Tutoring System, Software Engineering, Help-System, Java programming, Programming styles, personalised and customised help