9
$\begingroup$

Being a professional software engineer, i have seen a lot of really interesting technical work, many of it is published as technical articles in technical journals and websites like http://www.drdobbs.com and http://www.codeproject.com and many others, here is a few examples in some areas of interest:

Algorithm engineering and implementation:

Data structures:

Text search:

Object oriented and generic programming:

Database optimization:

Real time processing:

This level of work is considered highly innovative from the technical/engineering point of view.

I wonder if any technical work can be considered an academic achievement and can be deployed in research papers as well as technical articles.

In other words: When can a technical article be published as a research paper, in which areas? any examples?

Thanks

$\endgroup$
4
  • 7
    $\begingroup$ Could you say what your interaction with research papers has been so far? Have you read any? $\endgroup$
    – Vijay D
    Nov 16, 2013 at 21:35
  • 3
    $\begingroup$ Also, why publish them as research papers? If your community already reads those online venues and if you get personal benefit from papers then why publish academic papers? You already reach your intended audience. Research papers are no different than any other form of communication, they are just a way to share knowledge, and we have to remember that our goal as scientists isn't to write papers but to generate and share knowledge. $\endgroup$ Nov 16, 2013 at 22:48
  • 1
    $\begingroup$ Well, The main reason to ask my question is that i find scientific method really interesting and i wonder how i can contribute to science as well as technology/engineering, Another minor reason is to distinguish what is considered a scientific achievement from what is not. $\endgroup$ Nov 16, 2013 at 23:20
  • $\begingroup$ occasionally important results are published as technical papers (as in industrial R&D labs), more in the past, it seems now maybe the trend is more toward scientific papers, maybe partly due to a decrease in industrial R&D labs.... $\endgroup$
    – vzn
    Nov 17, 2013 at 15:54

1 Answer 1

8
$\begingroup$

Thanks for the question; I had similar questions few years ago, before starting in research (I'm not necessarily assuming that's your case).

I've looked at a couple of the links, and they don't really look like research papers in form; I mostly can't really tell if their technical content could be made into a paper because I'm not an expert in the field, but I'm guessing "no". If you compare your links to a paper, you'll notice several differences. But highlighting them and their rationale might help.

The key concepts are:

  1. Scientists learn new knowledge for humankind, and then share what they learn through research papers.
  2. Moreover, we live in information overload, so readers only have time for the most important knowledge.

As a consequence, research papers have some key features.

  1. A research paper contributes new knowledge in a field (novelty). To convince readers of this new information, it provides appropriate evidence to argue for it. What's appropriate evidence varies depending on the field and on what the paper claims.
  2. In fact, since a paper must be novel, it must also argue that its content is novel (though that's done somewhat implicitly). That's done by comparing to related works in the field; if another work looks similar to a submission, the author must explain what's the novelty in his work. Moreover, no paper is entirely novel, so a paper must also say what was already known by citing its invention (except for things that are standard). As a rule, technical articles don't do that; they typically don't claim to be novel. Sometimes the author invented the technique by himself, but it was already known.
  3. Good call for papers mention the criteria for judging papers — this one mentions novelty, importance, evidence and clarity.
  4. A good research paper should also contribute knowledge that is "interesting", and explain it clearly to the reader. What's interesting varies according to the publication venue.
  5. In fact, my favorite intro to research in book form is The craft of research, though what I've read is parts of this condensed version. I'd recommend to start reading it from the Amazon preview.

Example

The above is not really specific to computer science, so let's have an example. Since most of your examples are in algorithms, I'll try to show what the above means for an algorithm paper (even though I'm not an expert in that area, so I hope I won't get it too wrong). Such a paper might claim that some algorithm is fast, and prove that through complexity analysis or experiments. That paper better claim that the algorithm computes something useful (otherwise, who cares that it's fast?), often because it's correct; appropriate evidence is then a correctness proof. Moreover, this algorithm probably reuses ideas already known to experts (which should be cited), but has some key differences from any similar works (which should also be cited) — maybe it has the same complexity but is more general, or is simpler to implement, or has any other virtues that experts have agreed are interesting.

By contrast, a paper on, say, programming languages would have different claims and different evidence, but would still follow the same basic structure.

$\endgroup$
2
  • $\begingroup$ readers only have time for the most important knowledge. Probably true, though some wandering can help. But the concept of most important knowledge is defined differently for each person, depending on what she is doing, what she is interested in, the way she thinks, etc. $\endgroup$
    – babou
    Jan 9, 2015 at 20:40
  • $\begingroup$ @babou I don't imply "most directly applicable" or sth. such, though I probably should make that more explicit. $\endgroup$ Jan 12, 2015 at 8:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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