A Pragmatic Alternative to Bootstrap the Technical Contributions of a CS Project

Posted on September 3, 2014 by ivotron

In my last entry I talked about how a CS academic project can be conceptualized at many levels:

  1. Project (Goal)
  2. Sub-project (Purpose)
  3. Paper (Outcome)
  4. Technical sub-components (Activities)

Personally, I like to think in higher-level terms (1,2) but it eventually comes the time when we need to think in low-level terms (like writing an article). One alternative that I find works well is the following:

  1. State a level-2 research question to look an answer for.
  2. Determine the technical topics (CS sub-fields) that the question corresponds to.
  3. Search related work in those areas. Ask people for domain-specific keywords.
  4. Think what peculiarities your stance has w.r.t. the state-of-the-art and what insights can you obtain thanks to your distinct point of view.
  5. Design and carry-out simple experiments that demonstrate your new/enhanced technique or method w.r.t. existing work.

As one loops over this 2-3 times and document as you go, the content of an article will start to take form.