How to write a Survey in months?
A note aims to inspire those writing a survey for the first time.
First phase [collection and outline]
Almost six months ago, my advisors suggested that we write a survey in this field. Here, this note comes.
At first, I conducted a literature review and collected papers in this field to understand the volume of existing work. Following my advisor’s suggestion, I created a table to record the papers. Don’t worry about just few papers—over time, they will accumulate linearly.
I tried to classify the papers, but it was difficult to establish a clear classification system. Then, I quited, and drafted an outline for the survey architecture. The most efficient way to draft an outline is to find a reference.
Second phase [add, add, and add]
In the beginning, it was very hard for me to write anything meaningful. I have no idea where to start. Then, my advisor suggested, “Maybe you can start by adding one paper you’ve read.” Don’t worry about the connections, just add it.
So, we set a goal that every week, read, summarize, and add around three papers in this area. This is a long process, please hang in there, and be regular. You will have fun.
Third phase [draft]
At this stage, the survey still couldn’t be considered a proper article because there was no connection between different parts, especially between papers.
Luckily, you will have developed some ideas on how to make a classification system, and how to organize each parts. It’s amazing, but it’s the truth.
I first created a relatively mature classification system. Actually, I have modified several version before settling on one.
Fourth phase [polish]
First of all, the biggest problem lies in organizing different sections. If you jump straight into content without transition sentences, the flow will feel unnatural. Take some time in organization the transition sentence.
Second, after proposing a problem, always provide corresponding solutions or discuss existing progress.
Finally, a good visualization or table can leave a strong impression.
Tips
- Make a table about paper lists
- Set a regular writing habit
- Make connection
- Be careful of problem and solution sentence
- Good visualization
- Keep pace with the latest
Reference
Grounding Creativity in Physics: A Brief Survey of Physical Priors in AIGC
See you soon👋!