Ask HN: Struggling to find a Computer Science thesis topic

I have only abstract advice, but it's important advice. The thing that a lot of people miss when choosing a topic is that it has to be not just something that you're excited about working on but also something that you love reading about. Because situating your work in terms of the existing literature is a very large part of a research degree (and also will help you do good work, presumably).

Have you talked to your advisor about potential topics? The scope for an MS thesis is more like a large project (vs a PhD thesis), and it may make sense to ask your advisor about appropriately scoped projects/ideas that they have that are related to areas you’re interested in. It can be helpful to have your own ideas, but since they will have the most context and be the final arbiters of approval (if your school works like mine did), I would recommend that as a first step.

I would also say it depends on your interest and knowledge level. I would assume that before the MS, you haven’t yet approached the edge of knowledge to know what kind of work there even is to do (https://matt.might.net/articles/phd-school-in-pictures/). If you’re doing a research-based MS, that’s where the seminar classes, the papers, and collaborating on research can help, getting you closer to the edge to learn more about what’s going on there.

Sub modular functions