CSE 444

Research paper information

What was in the syllabus?

First-day information: lecture-A, slide 11.

Topic selection: Pick a topic and clear it with me. Please give me your proposal in writing, with your name and e-mail, that day's date, a title, and a sentence or two of description. By the time you make your proposal, you should have one or two sources already; please list them, too (no special format yet). The proposal can be handwritten the first time around, but by the second time should be word-processed. I'll let you know within a day whether it's OK to proceed with, or whether we need to talk further. You can also run them by Jake. Each person should have a different topic, remember.

Make the topic as specific as possible. It can cover a commercial product, but should not be a blurb ("IBM's new release of DB2 buries the competition"). Avoid application case-studies ("Federal Express's Database System") unless there's something really unusual about it. It could also be an academic research topic. I'll be especially impressed if your proposal involves topics related to other areas of CS.

BAD: Bull's IDS-II DBMS (too big)

BAD: Concurrency control (covered in the course, even though lightly)

MUCH BETTER: Concurrency control in Bull's IDS-II DBMS 

YOU have to find and select a topic. I won't do it for you! But if you're stuck, come to me and we'll talk. I realize it's hard when the field is completely new to you.

What if "there's not enough information available on my topic"? Then don't pick that topic! That's why research during the topic selection phase is so important.

I can get really picky, so factor a couple of slack days into your topic selection schedule. I won't approve a topic simply because the deadline has arrived (or past)!

How to find a topic? Some possibilities:

 The library catalog

 Periodical indexes (ask a reference librarian!)

 Leafing through recent issues of appropriate magazines and journals (BYTE, Datamation, Computerworld, ACM Trans. on DB, etc.)

 Internet browsing and searching

Be careful not to rely too much on an Internet search. It's not true that all information in the world is on-line, and even less true that what is on-line is well indexed.

The time you spend searching for a topic can really be time well-spent. During the process, you might end up accumulating all the information you need to go to the next step, which is....

Paper draft. Please prepare and print out a full draft of the paper, word-processed, including a complete list of sources (bibliography). I will be looking for organization, evidence of understanding, sufficiency of sources, etc.; looking less at specific technical content; and very little at format (except that format often reflects organization). I will also look at a portion of the paper for writing mechanics, so from the very beginning you do as well as possible with spelling, grammar, sentence structure, clarity of expression, etc. The draft should be word processed. The earlier you give it to me, the earlier I can give it back with some comments, so that you can perfect your....

Final version. The final version should be submitted both in hardcopy and in electronic format, either HTML or Word (PC compatible .doc file). I will grade on writing, both mechanics and style, as well as contents, quality of sources, organization, format, and general interestingness. More information later on these topics.

Length. That's the question on everyone's mind! This is only a six-week project, even counting the first week of class, and it will overlap other assignments. Length per se is not the issue, but what you have to say and how many pages that takes. If you do enough research, you will probably have material for quite a long paper, so the problem should be how to organize and trim it, not how to pad it out. If your final paper is really short, it may also indicate insufficient research, or an organizational problem, for example it contains insufficient background and explanatory material.

Schedule and grading breakdown (on a 16 point scale). Each step is separate, e.g., you can't make up in the final version points you lost on draft or topic selection. Lateness at any step will cause automatic point losses.

Topic selection (4): Completed (approved and word-processed topic turned in) Monday Oct. 13

Complete draft turned in (5): Wednesday Oct. 29

Final version (7): Monday Nov. 10

Presentations based on paper. Authors of a few papers will be invited to present them in class. This will have mostly to do with how the topic fits in with the course. You can't improve your paper grade by making a presentation. But there might be some tiny course grade incentive (like counting it as an additional perfect homework or something). There will be relatively few of these opportunities, unfortunately, so don't count on it as a way to improve your course grade.

Papers posted on course Web. Students who wish to can have their final papers (without grades or instructor comments) linked from the course Web.

Help along the way. The UW College of Engineering has a Writing Center precisely to help students improve their writing. Check it out!