Self-Evaluation for Spring 2012

1. When tackling a huge challenging project, figure out the primary conflict or core issue at the current stage, then solve it immediately and efficently. Hopefully, progress can be achived step by step.

2. Assume the audience does not know anything about your work and explain the intuition and general idea before getting into the details. Always show simple examples to illustrate the basic idea during a presentation. If possible, use a consisntent set of examples from the beginning to the end in a paper or presentation.

3. Do explain things in the slide in a presentation, do not put anything that you feel unclear about in the slide.

4. Describe your idea and design with motivation to attract others' interest in your work, for example, state the challenges or difficulties, then illustrate how your method solves the problem.

5. Come up with clues or solutions based on experimental results and observations.

6. Argue or describe in a logical way. Be critical and self-suspicious to your own argument. Do not be naively persuaded by your own argument.

7. Use experimental data to analyze your evaluation results. Interpret your evaluation findings based on quantitive analysis.

8. Do not put explanations or arguments without any solid or tangible evidence in a paper. It will only lower the quality of the paper.

9. It is as important to discover repeatable or reproducible empirical results as to propose solutions to some challenging problems in scientific research. Similarly, it is equally important to analyze the challenges of some problems and prove that they are indeed hard to solve.

10. When explaining your works' superiority or novelty, do not demote others' work. Instead, refer and acknowledge others' contributions, then address the limitations of the current approach.

11. Tell your supervisor what problems you have solved or what progress you have achieved, instead of what you have done. They are actully less interested in the latter.

12. Make your experiment management scripted to reduce redundant labor and human errors.

13. An abstract is a highlight of your contributions. Explain what significant problems you have solved with the help of what you have developed, and highlight the evaluation of your solution.

14. Making a daily or even hourly schedule is essential for fast-pacing research work and software developing.