7

I have an academic problem that I'm very interested in and want to explore the related research fields. Of course no researcher knows everything going on in a field but I feel I know far too less to go for a PhD right now. I'm in the last year of my undergraduate program right now.

What I did so far is...

  • Read a lot of papers. They build on a large corpus of knowledge so before understanding the theoretical details I have to read all those somewhat related publications.
  • Read advances papers for better understanding. This works quite well.
  • Organize links to all interesting publications and other material in a document. Not sure if it's really useful though. I haven't had time to read things from there I didn't remember anyway yet.
  • Start implementing the ideas and algorithms as computer programs. Doing so is good to validate my understanding but also takes a lot of time and it would be good to get some help when I'm stuck.
  • Join mailing lists and groups and talk to researchers. It's more networking than information.
  • Take somewhat related classes at other universities since there is not much at mine. However, there are no whole courses on the problem I'm interested in since it's too specific.

This all is a lot of fun. It also obviously takes a considerable amount of time. Do you have any advice to more efficiently organize and catch up with the current state of my academic field of interest?

  • 4
    If it is a reasonably well established field, read books first. Doing so uses experts in the field to organize the foundation information for you. – Patricia Shanahan Oct 14 '15 at 22:30
  • 1
    You need to focus. Doing that for a wide area is impossible. You need to focus on a small area and then read everything about it. The "Start implementing the ideas and algorithms as computer programs." is impossible. For CS it will take you as much as six-months to implement one complex algorithm of a single paper. – Alexandros Oct 15 '15 at 9:10
  • 2
    Nice list. When you talk or email with researchers in the area, ask them for their top ten, or top three, or whatever, favorite books/papers. – aparente001 Oct 17 '15 at 2:36
  • 1
    @aparente001 I like that idea! – danijar Oct 17 '15 at 11:40
7

the following points are based on my experience, on how to start learning about a research topic from scratch:

  1. Common Knowledge: First thing you do is to learn the common knowledge. This is the basic that makes sense of the whole field and logically true. For this, you can read general available resources, such as books or Wikipedia. For example, if I want to start to do a research on operating systems in a distributed setting, I go and get a book about Linux, and a general book about distributed computing. Then moving on to computer cluster for example or something else, I see what my interests and curiosity takes me.

    Note on 1: Don't think you are not reading research so you are not progressing. Common knowledge now was a research topic of the past, and you must learn them pretty well.

  2. Earliest Resources(s) You Can Find: This might take a little time, but you need to find the first resources to the field. You should see how and for what this field is started.

  3. Important Papers of the Last 20 Years: Find the 'top 20' papers of the last 20 years. See what they cited and what they contributed.

    Note on 3: The last step will definitely take you for a ride for some time. And this is the whole point. You need to learn how to read and write research in a field during a Ph.D. and therefore this step is very important.

  4. Day to Day Updates: You can join different mailing lists of different research topics, and also you can use online services, that let you know automatically via email, based on your specified keyword(s).

Conclusion: Learning the research topic and its history takes time; however it is essential to make you a good researcher who can contribute in the field. These things can be achieved by the following mentioned points.

| improve this answer | |
5

It depends a bit on the breadth and timeliness of the problem you have in mind, but it may be worth considering writing a survey paper for an undergraduate journal. It could force you to focus the task a little. Moreover, you would simultaneously develop research skills and publication experience, which may be even more important for PhD admissions than your familiarity with the specific related work.

For this to be a reasonable option, I would suggest verifying the following:

  • That there is a professor at your university who may be willing to offer some guidance (this is quite likely if you mention your intention)
  • That there is a suitable undergraduate journal for the problem you have in mind (your university may even publish its own undergraduate journal)
  • That you can narrow the scope of the problem to something that you could reasonably address in twenty-five pages or less
  • That the problem is still of interest to the research community
  • That you first have the basics of the field mastered (but that is quite likely if you are already in your senior undergraduate year)
| improve this answer | |
  • That's a good idea. Explaining things to others is a good way of learning. I study computer science but is no professor at my university who does work related to my interest. Thanks for the suggested steps! – danijar Oct 15 '15 at 19:03
4

Regarding keeping up to date: one thing I found useful is to have a google scholar account where I can get alerts whenever new paper in areas I specified are published or are uploaded to some kind of archive. This does not of course mean you have access to all new material as it is limited by how easily google scholar can find this material, But I found it the most effective for two reasons:

  1. It scours everywhere for material, so I am not biased by few pre-determined sources (as opposed to for example only looking at articles from a specified journal homepage)
  2. Google is the best search engine yet in terms of providing a complete list of citations (although the citations do not necessarily come from published articles only) so you can evaluate the impact of the paper in a way.

[Disclaimer] mentioning google here as 'the best' is based on just my experience, I once came across a published work regarding research impact and citation accounting/quantifying methods, will post here if I find the link again .

| improve this answer | |
  • 2
    Note that Google Scholar alerts are based on "new to Google Scholar", rather than "newly published". I have a general alert set up for my institution's affiliation (it's a good way to catch obscure papers) and I do get a fair amount of notifications for recently included papers from as far back as the 1970s... plus a lot of odd material that gets indexed, including blogs, newsletters, and in one remarkable case a Spanish-language thriller novel. – Andrew Oct 15 '15 at 11:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.