Is it possible to do CS independent research out of school ?

Question

I am a undergraduate student in Computer Science and wants to pursue a PhD. Since I don't have a good GPA, I've heard that I need to get some research experience to enhance the chance of getting admission.

However, Due to personal reasons, I'm taking an year off school to go to Substitute Military Service. Because of the nature of the Job, I might have plenty of free time for self study. So I'm thinking of doing research by myself. Do you think it is possible? If so, how should I do it?

I've thought of some topics I'm interested in. One is Image inpainting , another is Physics based animation. Though I'm still not sure how to start, since I don't have any research experience before.

Thanks for advice.

Answer 1

Typically, research implies that your work will be peer reviewed. This means that you'll publish your work in a public forum, either in a journal or at a conference. Personal projects are worthwhile, particularly during under-graduate studies, because they are one way to impress academics into accepting you into their research group.

But honestly, without peer-reviewed research output, I can't see it making that much of a difference.

Answer 2

I assume you wish to do the PhD in CS, not another discipline.

You might consider becoming a contributor to an Open Source project in an area which interests you. In this way you will get a feel for what level your skills are. You will also be part of a community. I have known some very good contributors who did not have good formal qualifications.

However you will have to work hard and earn the respect of the others. To do that you will have to read a lot - you cannot expect others to help educate you.

And to be very clear it is very unlikely that this will lead to publishable research. It does occasionally happen but it's very rare. And success here will probably not make any difference to your chance of getting accepted.

Answer 3

I suggest you enrol in some Computer Science Distance Classes/Courses at a reputable College/University.

Here in Sweden a lot of distance learning is usually done at a half or quarter the normal pace to make it easier to do alongside a job or studying something else.

Answer 4

I can talk about my own experience. In the past I have spent long periods of time doing research in software engineering "on the side", either in my spare time because my job was not research-related, or while I was unemployed. It is not easy but it can be done.

In my experience, you will need two things. First of all, you'll need to get your results published. As Andrew Walker said, without published results, research is of little value. Determining the appropriate venues (journals, trade magazines, conferences) where to submit your work is a tricky task to do, and given the vast diversity of fields in software engineering and computer science that exist today, you'll need to focus on one field before you start thinking on where to publish.

The second (and foremost) important thing you will need is to learn how to research. This means learning how to perform a literature search, how to defend an argument, how to structure a technical text, how to present a result, how to draw conclusions from data. All this may sond trivial, but it's not. Mastering these "research techniques" is what differentiates a researcher from an amateur. Usually, a PhD will polish these skills for you, but advancing them now if you have the opportunity would be great in my view.

Happy to continue the chat. And good luck!

Answer 5

While it is technically possible, I would highly recommend against it.

First, if you are interested in research on human factors or software engineering (or just do user testing), you might need to have legal/ethical permission to do human-based subject research. In some countries this is limited to those who receive permission from an Institutional review board. While it is meant primarily for medicine and psychology (i.e., no Dr. evil who's experimenting on people in his secret lab), it applies to all other forms of research. If you don't have approval, you might be breaking the law or having problems publishing.

Second, research takes a lot of time and has to get published as papers. While some conferences do dual-blind reviews, many still expect you to be affiliated somewhere. If you do get accepted, you also need to pay for your trip :)

Third, it is good to have peers who review your work and are up to date on research before you spend a lot of time on something.

Fourth, you might need special equipment or labs (e.g., if you're doing networking research and need special setups)

Also, if you are currently employed somewhere, they may have rules about conducting research even in your spare time.

Answer 6

For pure knowledge sake check out http://academicearth.org/. This site offers lectures in alot of subjects from many ivy league colleges. There is a whole lot of info here.