### Present Remotely

Send the link below via email or IM

• Invited audience members will follow you as you navigate and present
• People invited to a presentation do not need a Prezi account
• This link expires 10 minutes after you close the presentation

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.

# How To Get Your Research Paper Accepted?

Presentation for the Imperial College MRes/PhD Research Seminar
by

## Peter Pietzuch

on 17 January 2013

Report abuse

#### Transcript of How To Get Your Research Paper Accepted?

How To Get Your Research Paper Accepted? Rule 0010 About Me Rule 0001 Peter Pietzuch
Department of Computing
Imperial College London
<prp@doc.ic.ac.uk> Resources Background in distributed systems, networking, databases and software systems research
Many rejected papers over the years (and some accepted ;-)
Serving on many other systems and database PCs
Disclaimer
It's specific to my research area...
These are my own thoughts, others may disagree
You don't get you money back if your paper gets rejected Make sure that
there is a problem Who Cares? ￼￼Scientific writing
One of the most important things that you learn as part of your PhD
In your career, you’ll be writing emails, research reports, technical specifications, whitepapers, patents, blogs, client presentations, ...
Crucial to communicate your ideas to other people—only that much can be done orally

You want your papers to get rejected because of the science and not the presentation (eh, wait...)
Aim to always get the comment: “this is a well-written paper” This Talk Cut to the Chase Keep the reader interested
Don't just report what was done
CS research is not maths or physics
Always explain design decisions
Keep the momentum in the paper State your Contributions Make it easy for readers/reviewers to understand the novelty
Don't claim the wrong things
Relate them to the state-of-the-art Rule 0011 KISS Rule 0100 Make the research accessible
You don’t want the reader to work hard to understand your work
Give the reader the bigger picture first
Figures and examples are your friends Writing Style Matters Use present tense
Gives impression that your work is valid in the present
Don’t use past tense or future tense
Use active voice
Makes it clear in terms of who does what
Passive voice keeps the subject hidden
Avoid any repetition
Feels condescending (goldfish memory?)
But restate key points/message with more detail Rule 0101 Use Scientific Language Use precise and formal language—every single word should add to the meaning
Readers notice hand-wavy language—you can’t sweep issues under the carpet
Always define terms, abbreviations and variables before using them
Be consistent in the choice of words: “the system” vs. “the prototype”
Don’t use words that convey your judgement:
very, bad, poor, fortunately, unusually, clearly, excitingly, it should be noted, . . .
Don’t use "weasel" words that are vague or ambiguous:
rather, arguably, relatively, often, probably, some people, many, in most respects, . . .
Figures and figures should be self-contained on their own Rule 0110 Use a Sensible Structure Compare to the State-of-the-Art Discuss it in relation with your own work
Don’t just enumerate, always compare
Don’t be dismissive
You're not selling a used car
Include a discussion section
Removes negative points from the reviews Rule 1001 Think about the Evaluation Choice of experiments matters
Think carefully about what the experiments are supposed to show
What questions will the reader have?
Discuss all results and draw relevant conclusions

Evaluation template
1 Evaluation goals
2 Evaluation methodology
- Overview of experiments
- Evaluation metrics
3 Experimental set-up
4 Experiments
5 Discussion of key insights Rule 1010 ￼

￼￼ 1 Abstract
2 Introduction
3 Background/Related Work
4 The Meat
5 Evaluation
6 (Related Work)
7 Future Work & Conclusions
It’s fine to deviate a little bit based on your material

Example structure
Motivate problem
- The Internet crashes without wibble
- Billions of dollars are lost and we all die
Key idea
- Implement wibble as part of every middleware
- Can be done thorugh RPC
Detailed contributions
- Describe the wibble algorithm
- Explain interactions with RPC Pay Attention to Detail Typos, layout flaws, small mistakes reduce confidence in the science
We're human and judge presentation
Run your paper through a spell-checker Rule 1011 “The Elements of Style”, William Strunk Jr. and E.B. White. Macmillan Publishing Co., New York, 1979. http://www.bartleby.com/141/strunk5.html

“Writing Technical Articles”, Henning Schulzrinne http://www.cs.columbia.edu/~hgs/etc/writing-style.html

“Tips for Writing Technical Papers”, Jennifer Widom, January 2006 http://infolab.stanford.edu/~widom/paper-writing.html

Justin Zobel, "Writing for Computer Science", Springer, 2004 1100 Rules that are Easy to Follow Start with a clear and compelling problem statement
Motivate the problem
Show that it is challenging
Explain why you did this research Questions? If you have any feedback
about missing rules, better rules or wrong rules, send an email to
<prp@doc.ic.ac.uk> Quality of writing improves with practice,
ie write and get feedback Paragraphs are essential to structure your material—use them!

Each paragraph should express a single thought, point, argument, ...
First sentence of a paragraph is important:
Should contain the message of the paragraph or summarise it
Make sure your paragraphs are balanced in length
A single paragraph spanning an entire columns is not a good idea! Sentences
Control the length and complexity of your sentences
Don’t complicate sentences needlessly—shorter is better Long sentences are hard to parse
Avoid breaking the flow by using parenthesis or footnotes
A single sentence spanning an entire paragraph is not a good idea! Rule 1100 Make your Layout Choices Consistent! LaTeX is your friend
Will result in professional consistent layout
Learn it!
LaTeX pet peeves...
Use protected spaces (~) to avoid bad line breaks
Figure~\ref{fig:arch}, Section~\ref{sec:intro}, Wibble~\cite{wibble11}, . . .
4~MB/s, 10~nodes, . . .
Type-setting multi-character variable names in math mode
$varname$ typesets as "v a r ..."
Instead use $\mbox{varname}$
Full transcript