Center for Analysis and Design of Intelligent Agents

T-720-ATAI-2021: 'Mock' Abstracts & 'Final' Abstracts

Purpose	The purpose of writing an abstract for a scientific paper is to serve as a guide to your readers.
What an abstract should say	Having read a paper's title and the abstract, a reader should have a crystal clear idea about: * What key question(s) are addressed in it * How they were addressed before you did your work * Why that was that insufficient to address them * How you are addressing them * What your main methods were * What your main conclusions were

What is a 'Mock Abstract'?	A 'mock abstract' is an abstract that is written before the paper is written.
What is its purpose?	To help guide you during the writing the paper.
How can it be written before the paper is ready?	It can't. Not really. Which is why it will be rather “rough” the in its first few iterations. But taking your best shot at a mock abstract will be tremendously useful (see below).
How often do I have to rewrite the mock abstract?	Until it's as perfect as you have time for. You are very likely to go back and re-write your abstract many times during the writing of a paper. Therefore, it slowly evolves from 'mock' to 'final'.
Why do it before the paper?	Because it helps you tremendously stay on track when writing. Which means it speeds up writing. It also helps it be more clear.
Does it have to be perfect?	No. That is impossible. Besides abstracts are never perfect, ever. They can always be improved.
What does this imply?	The abstract should be the first and last thing you work on for a paper.

Each item below should take only 1-2 sentences in your abstract (unless otherwise noted)
Outer context	A short introduction to the field or context of the research problem addressed. This places the research topic within the larger scientific context. Go for “the shortest introduction possible”.
Research problem / question	A short introduction to the research problem. Explain the problem you are addressing, in as simple words as possible. Make sure you describe it at the right level of detail. Not too general - then you are repeating the above point; not too specific - then you leave out some context that is necessary for some of your readership. “The shortest problem presentation possible” (max 4 sentences).
Key Motivation	Why this work was/is worth doing. Make sure the justification is scientific (not personal! - stick to verifiable facts, in the context of scientific progress).
Prior attempts at addressing it	How has this issue been addressed before?
Why prior attempts are not enough	How do prior attempts fail at addressing the issue?
Method	How did you do perform this work? (max 4 sentences at most).
Results	What were your results? (max 4 sentences).
(Conclusion)	Sometimes done to emphasize the main message or conclusion of the work described in the paper.
If you succeed…	…in following this guideline, you have a guaranteed great, readable, informative abstract that is neither longer nor shorter than it needs to ber. In other words, the perfect abstract.