My Reflections on My Boring Research Achievement

so far, all my research achievements were made through the same recipe as follows:

  1. Choose a Problem The first thing is to choose an NLP task to work on. The famous saying is "Choice is more important than effort". However, if I do not have much choice (e.g.. being asked to solve a fund-related/project-driven problem), that is also good. The reason is that then I have to stay focused, and persistence warrants achievement in any problem.

  2. Read Papers An exhaustive literature review makes me really understand what to do. Usually, I will start with all *ACL papers that are published in the last three years on the same topic. The reading list will keep growing as the reading progresses because more and more papers will be found in the Related Work sections.

  3. Brainstorm The next thing is to brainstorm my own solution for the chosen problem. A worth trying idea for me must fulfill the follows:

    • Be different (unique). This is always my top1 concern (my research taste).
    • Fix something (find and attack the weaknesses of previous research). So it is easy to persuade some reviewers later.

  4. Experiment and Refine In most cases, my initial ideas would not work out. However, after some experiments and refinement iterations, they would eventually evolve into good ones.

  5. Write and Submit To be honest, and sad to say, publishing is always the core motivation for my above endeavors.

The reason for me to write this up is that I get bored and lose my passion for research. I do not feel excited about what I do, and sometimes I feel embarrassed to talk about my research. It is pretty boring to repeat the above process, but I do not know how to make a difference.

I have to admit that:

What will I feel excited to do?

What makes an exciting research to me?