#笔记# How to do good research?

Speaker: Michael Fang (University of Florida)

1. How do you train your vision?

• Through the process of doing research
• Summarizing all you have done and putting it in the perspective (from a tree to a forest)
• Thinking philosophically (methodologies and tools)
• Generalizing and checking whether what you invent can be applied to other fields
• Learning from the master !

2. How to choose a good problem?

• Make sure that the problem is important and timely
• Articulate that the problem has practical relevance
• Demonstrate that your problem is of great interest to its community
• Ensure that the problem is challenging
• Predict that a possible solution to the problem will generate significant impact

3. The solution of a problem does not make you great, the problem itself does

• The solution gives you the feeling of ending, while the problem indicates the beginning!
• Problems can always motivate you to think, to create, and to search for solutions
• Only when your problems are important can your solutions be recognized and appreciated more
• Good problems tend to lead to many innovative tools useful for other problems
• Frequently check whether you are working on important problems!

4. How to write a good paper?

• It is NOT your English problem! Technical flow is the key!
  • Be aware of the logical flow: organization may give better impressions to the receivers
• Know the culture of the publications(format, requirements, acceptance rate, people or community who will review your paper etc.)
• Minimize the easy mistakes reviewers may use to write quick reviews! (Everybody writes and nobody reads!)
• Minimize the typos and grammatical errors!
• Write a good abstract, a good on should tell the real contribution of your work in this paper: this is crucial, you must impress the reviewers at the first sight!
• Write a good introduction: important! Many reviews only read this part to get the idea or find your faults
• Use a comprehensive and up-to-date reference list: missing one crucial paper may lead to rejection
• Do not be confrontational when replying or responding to editors and reviewers

5. How to write abstract?

• What is the problem to be addressed?
• How important is it?
• What is the current status?
• What are the problems in the current solutions, if any?
• What is the intuitive idea of your solution to address these problems?
• What is your finding? What do you conclude from your study?

6. How to write introduction?

• What is the problem to be addressed? If the problem is well known, you can skip this, but it is always good to have this
• Expand on how important the problem is
• Give a comprehensive survey on the problem research
  • Comment on current solutions and point out the strength and weakness
• Identify the problems to be addressed and offer the intuitive solutions with intuitive ideas
  • Find a simple yet illustrative example to demonstrate the interesting intuitive ideas
• Summarize the main finding and conclusions

7. When do you feel you can graduate?

• What does a Ph.D degree actually require?
  • Fundamental training: course work
  • Fundamental skill: the Ph.D training process
  • Ph.D dissertation: your work
• To be a good Ph.D, you have to have the feeling of being a PH.D
• Ph.D is Doctor of Philosophy, you need to have the feel of quantum jump during your Ph.D study process
• You must have the feeling that you actually accomplish a little and need to push a little further
• "When you feel you do not know anything they give you a Ph.D"
• You need to have both the breadth and depth in your own field and you must be an expert on your topic
• You must have the skill set of problem solving
• You must have had the moment or feeling or sensation of "quantum leap" during your Ph.D study!