CSE 585 HW#4: Improved Research Proposal (+ Document)

Written Turnin Due: November 15, 11:59pm, turnin: on Canvas.
OPTIONAL Presentation: November 15 or 17, ~10 minutes each
HW#4 will be graded on the written document you turn in.
If you re-give your presentation, it may retroactively add points to your
HW#3 grade, but only if your presentation is better than last time.

In this homework you should improve your HW#3 proposal (or develop a superior alternative) in one or more of the following ways. The goal is to model a mini MS-thesis-proposal type of document. Note that working scientists have to write proposals all the time.

Make your research question(s) more explicit

Many of your HW#3's discussed a technical topic without explaining clearly a question/problem that is unsolved that you might hope to solve.

Relate your research question(s) more strongly to the related work

Many of the HW#3's did not make clear the relationship of the research question to a rich body of related work.

Propose more concretely how it might be answered/solved

Have an explicit section on experiments/methodology, explain how the data will be acquired and analyzed.

Lose all self-awareness

While we often want our computers to gain self-awareness, research proposals are not about you, they are about the field under study.

Add depth

Go ahead and write a written document for your proposal. Please dig in deeper into your topic and its related work, and produce a more thoughtful proposal.

Acknowledge and respond

A proposal does not have to present a finished research result for which all counterarguments have to be addressed. However, you should consider how a polite skeptic who doesn't believe or agree with your positions might challenge the work you are proposing.

Budget

Include a brief estimation of time frame, labor (time) and space costs it would take to complete this research and answer your research question. Note that some research is much less onerous than others and this is OK. Do not propose something that would take > 2 years.

Format

Once again, please learn and use an ACM legacy two-column conference paper format. Some of you have not approximated it very closely thusfar.