But that paper proved teleportation?! (Part 2)

How to write a paper?

The best way to write a publication is:

1. get a pen and write the generic pieces of a paper".
           (my organization)
• intro
• related work
• pipeline/overview
• tech details
• results
• limitation/conclusion/future work
  
  (the organization on the slide)
• title & abstract
• authors
• intro
• related work
• methodology
• findings
• conclusion 
2. Read lots and lots of papers. Copy/borrow the structure of the papers you like and understand.

Parts of a paper

Title and abstract:
- Title should descriptive, include keywords, make it simple, don't use area-specific jargon, avoid typos because that makes the senior reviewers to decline reviewing your paper.

Authors:

• Be generous, invite people to be an author in your paper. It builds your network, ask people whether they contributed enough to be an author, ask explicitly. If someone asks you, it is a responsibility, you should feel confident that you have contributed enough, something you have verified and believed that it is true. Even if you did not do it, you are responsible of the typos, wrong things in the paper, and it is still you responsibility.
• As woman in cs, we might be not claiming what we did...
• How to be a good co-author? You have a team that worked on that. Do really explicit what you do, write down everything you did. Do bulleting!!! (like my advisor says...). You might be working 24 hours but other people may not.. Let others speak too, we are all collaborators, feel comfortable to ask, say, contribute.

Intro:

• What is the contribution?
• Be concrete, give example. Don't be abstract. Don't start in low level and go to the big piece. All these cute ideas that can't be squeezed in a paper, create a dead kittens file. Don't include, put that in the other file.

Related Work:

• Least liked part!!
• That's where you can show what you did is better in the area of your world! Put yourself in your area. You should be aware and thoughtful about who you cite and how you cite them. You don't want to be overly defensive, and don't make it an island in a huge ocean. There should be some similar stuff somewhere in the world. You don't want to cite all of the world, just cite them in most recent and useful manner.

Methodology:

• Tell people how you did what you did.
• How to reproduce, the exact approach
• Acknowledge limitations, frame the limitations.

Findings

• Clearly state what you observed, put numbers in the text..
• Try and pull all of your findings, numbers, etc. to tables, figures, etc. They should be standing alone. Somebody that is flipping through the paper should understand a paragraph alone, or a figure alone.
• Help the reader interpret the findings.

Conclusion

• Recap the abstract.
• A person can just read the abstract and conclusion, so make sure that you have a good summary.

Other concerns

• Ensure the proper layout
• Take some time that the paper is really on the product line
• Take professional help about English if needed
• Submit the paper on time and read the call for papers
• Don't ever plagiarize. Ask for permission. Cite and acknowledge. Don't take text directly from anywhere, you might forget to go and change.
• Submit the work you are proud of. The number of submissions to conferences is increasing highly, send it to colleagues if you need feedback.

How to deal with the reviews! (Especially reviewers...)

• Write a response, wait for a day to send.
• Some people are strong enough to read the reviews momentarily, but some are not. Just put it away, and come back to them when you are ready.
• Look for the things that keep coming up.
• If somebody says "I don't get it", just explain. Don't say it like "Oh I did say it in section 4 par 5...". Maybe you mis-communicate, if most of the reviewers do not get it.
• If you respond the reviewers, they will see the paper again, and somebody might be pissed off..

In case of Rejection:

• great papers sometimes get rejected
• 3 times is kind of enough to resend
• or consider a venue change

In case of Acceptance:

Camera ready:

• take the feedback into consideration, still
• share the paper, blog about it tweet about it , present the work
• leave the details in the paper