8

I am writing my first research paper in field of computer applications and artificial intelligence. I have following doubts, please answer them to help me out..

a) When explaining my research, do I need to include the whole code or just algorithms ?

b) Do I have to upload my code somewhere so that people can verify my research and give link of that code in my paper ?

c) I don't know any professor very well. Is it required to get it read by somebody because I am still an undergraduate ?

Any other suggestion from your experience ??

Improve this question
5
  • 3
    You are asking very important questions. However, for this site it's better when questions are separated. Could you separate a)b) from c)?
    – Piotr Migdal
    Commented Nov 22, 2012 at 23:42
  • OK, sure. For this site it is the best when questions are split, i.e. one question per one question thread. Otherwise it may be hard to answer, or one can know answer only to one question. Is such cases you edit your question post to contain one question and create another question post with the other one (e.g. even copying and pasting). But not worries, it didn't sound rude to me... rather - humorous. :)
    – Piotr Migdal
    Commented Nov 23, 2012 at 1:54
  • Ok, Got it. Next time I would take care about it.
    – user1756648
    Commented Nov 23, 2012 at 2:02
  • I am assuming this is for a conference or a joural? Or is it a final "paper" for a class?
    – dearN
    Commented Nov 23, 2012 at 19:38
  • @drN It is for research journal.
    – user1756648
    Commented Nov 24, 2012 at 15:07

2 Answers 2

Reset to default
7

  1. Definitely not the whole code.

  2. It is a conference-specific and depends on the conference area. If the paper is more data related (which is the case in Data Mining field for instance) you are required at least to comment on the data usage and from where you get it. If your paper is about a new web system usually referring to it in the text will be helpful.

  3. It is not required but highly recommended since they will be able to provide a constructive criticism for your work. You do not need to know them very well and usually they will be happy to help students.

Generally, students (specially undergraduates) publish papers with the help of a faculty member in the school. This said, there is nothing limit them to publish by their own. It is just more helpful to you to be attached to one of the faculty members and do research under his/her guidance.

Improve this answer
2
  • 1
    +1 I would recommend getting a supervisor, I could not imagine myself writing anything close to a good paper when I was an undergraduate. But that might be just me...
    – Paul Hiemstra
    Commented Nov 23, 2012 at 8:26
  • And the supervisor will pay the fees if they are any.
    – Zenon
    Commented Sep 28, 2013 at 19:47
4

a) When explaining my research, do I need to include the whole code or just algorithms ?

The complete verbose code wouldn't fit into the paper anyway, or would it?

Seriously, what you need to ensure is that

  1. you clearly communicate your ideas and the contribution to the state of the art you claim to have made;
  2. the evaluation that your proposals are feasible and really as good as you claim is plausibly reproducible. In the simplest form, that means that you describe all the conditions and steps which when repeated should yield the same results as you obtained.

b) Do I have to upload my code somewhere so that people can verify my research and give link of that code in my paper ?

You don't have to, most of the time it is not being done. However, myself, I strongly believe it is beneficial and I hope the field will turn more towards this practice. For ideas about this see also the Science code manifesto, as well as Matt Might's CRAPL license.

BTW, think the same about your datasets you used for the evaluation in your paperr.

c) I don't know any professor very well. Is it required to get it read by somebody because I am still an undergraduate ?

It is not required, but very much advisable. If you feel ideas are worth anything, try to approach somebody in the field (e.g., an author of one of the most prominent references your work builds upon) and give them time to digest. You might get some good feedback. But even if not, try to submit to a conference/workshop where it would fit, you'll get plenty of feedback in reviews.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .