The Data Mining Forum                             open-source data mining software data mining conferences Data Science for Social and Behavioral Analytics DSSBA 2022 data science journal
IMPORTANT: This is the old Data Mining forum.
I keep it online so that you can read the old messages.

Please post your new messages in the new forum: https://forum2.philippe-fournier-viger.com/index.php
 
Writing a Survey
Posted by: David
Date: December 09, 2018 11:17PM

Hello,

I just started to collect information to write a survey,

Could you please let me know the best way to write the survey?

Kind regards,
David

Options: ReplyQuote
Re: Writing a Survey
Date: December 16, 2018 12:25AM

Hi,

That is great. Writing a survey is a good thing to do. If you write the survey well, the paper can be cited by many people.

I think that the most important when writing a survey is to make sure sure that you understand the topic well. Your survey should provide a summary of the existing papers. Thus, you need to read many papers, to understand them, and then to extract what is important from these papers. But also, you should try to categorize the work described in these papers. For example if you do a survey about pattern mining algorithms, then you could categorize the algorithms according to some criteria like : breadth-first search vs depth-first search, parallel vs sequential algorithms, algorithms that generate candidate vs don't generate candidate, etc. Thus, in your survey, you can make a table that summarizes the papers and classify them according to your criteria into several categories. This will help the reader to understand the topic.

A big mistake that I see in some survey papers is to just list all the papers one after the others but to not to classify them into categories. This is not good. It shows that you have read all the papers but were unable to summarize and classify them.

Also, you may want to talk about the advantages and disadvantages of the different approaches.

In my survey papers, I like to talk about:
- why the problem is important
- if there exists some survey published by someone else, I explain why we need another survey. For example, other surveys on this topic are old, so we need to make a new one.
- Then I divide a section into some subsections to talk about the different papers. Each section is a subcategory of papers.
- Then, you can also talk about the research opportunities. What do you think are the current problems in that field? This is interesting
- You can also talk about software or implementations
- A conclusion.

Some other advices:
- You may take some good survey paper as example and try to replicate the structure. This will help you. But if you do that, choose a good survey published in a good journal as example.
- You may think about a journal where you will submit your survey, and use their format to prepare your paper.
- Always keep in my mind that you write the survey paper for someone that is not very familiar with your topic. It is a little bit like teaching... You need to organize the ideas in a way that make sense for the reader.

Hope this helps.

Best regards,

Options: ReplyQuote
Re: Writing a Survey
Posted by: David
Date: January 03, 2019 05:04AM

Very helpful

I read your surveys they are well categorized.

thanks for your help

Options: ReplyQuote
Re: Writing a Survey
Date: January 06, 2019 04:08AM

Thanks. ;-) Glad my surveys are useful.

Philippe

Options: ReplyQuote


This forum is powered by Phorum and provided by P. Fournier-Viger (© 2012).
Terms of use.