Thesis Proposal Defence

My Thesis Proposal Defence, the second major milestone of the PhD program at Dalhousie’s Faculty of Computer Science, will take place on Thursday December the 19th, 2024.

Title

Exploring Continuous Interactions for Generative Machine Learning

Abstract

This thesis proposal provides an overview of the research I have completed as part of my PhD program to date as well as highlights the planned research that will be completed during the remainder of the program. My research is positioned at the intersection of Human Computer Interaction and Machine Learning and Artificial Intelligence with a specific focus on helping end-users better utilize these emerging technologies. The overarching theme of my research is to explore how support tools that leverage generative technologies can be designed to support continuous interactions such that users can seamlessly integrate them as part of their unique and individual workflows in various domains. I propose to achieve this through several related projects that together will form my contribution to this field of research.

This thesis proposal document first provides an introduction to the research area as well as a discussion of my overarching research objectives, followed by four research questions that together serve as a guide to my research agenda as a whole. Next, I provide an introduction to the related literature, including on topics of HCI, Support Tools, Machine Learning, and DMIs. This wide range of related work is not only necessary but crucial for positioning my own work as a meaningful contribution to the field. Next, I propose a four-project approach to my thesis research, each of which building off one another in an effort to answer my overarching research questions. These projects include exploring how to create low-dimensional latent representations of synthesizer patches and how those representations can then be used to generate new patches, developing open source mapping middleware tools, leveraging generative models to better understand fuel consumption aboard small marine vessels, and the design and evaluation of a novel “agingDMI”. Each of these projects provide contributions to my research area, and together aim to provide generalizable insights that support the essence of my research objective.

Finally, in this proposal I will provide a thorough overview of the my proposed thesis’ structure, including proposed chapter outlines and summaries. Combined with a rundown of the research contributions each of the aforementioned projects make to the field, this thesis outline will also serve to justify the suitability of my research for the completion of the PhD program. In this document, I will also provide an overview of project specific results that have already bee completed, as well as any anticipated results for work planned for upcoming semesters. I will also briefly discuss how the results found in these projects can be generalized for other projects working on support tools with generative systems, as I expect these insights to be valuable to other researchers in the field. Along with these results, a timeline of proposed milestones is presented where I show the planned work required to complete my degree in a timely manner. Finally, I conclude this document with a brief summary of the work completed, the work that remains, and the overall plan for the completion of my PhD program.