Quality of the Question

• You really need a question that can be answered.
• You don't want a question like, "What is the best knowledge representation format?"
• This has a lot of problems.
• How do you measure best?
• How do you even enumerate all possible knowledge representation formats?
  
• On the other hand, you need a question that is interesting.
• A question like "Is XML a reasonable knowledge representation format?" is pretty weak.
• The answer is pretty obviously yes.
• Herb Simon paraphrase. If you're working on a hard problem, if you solve it, you actually get somewhere.
  
  
• How about, "Is XML a good knowledge representation format for John Deere for tractors?"
• Tip 1: you can easily transform a statement into a question.
• Tip 2: you can rewrite the question later.
• Tip 3: keep the question in mind while doing the rest of the research.