Dept. Of Mechanical and Industrial Engineering University of Illinois at UrbanaChampaign

1
Academic PublishingPurpose, Processes, Pitfalls,
Perks

• Andrew Alleyne
• MIE and CSL
• UIUC

2
Outline

• Purpose
• Why do we bother to publish?
• Processes
• What are the mechanisms associated with getting a
  paper published?
• Pitfalls
• How do people usually fall short in publishing?
• Perks
• Whats the upside?

3
Purpose

• Why do we publish?
• To avoid perishing??? No.
• Answer to be part of a community.
• The purpose of Research, particularly in
  academia, is to generate and develop ideas.
• But how do you know whether your idea is
  worthwhile?
• Youll never know just sitting in your office/lab.

4
Purpose

• As part of a larger academic community we have a
  responsibility to share our ideas with each
  other.
• This is how we compare ideas and find out which
  ones are worthwhile and which ones arent.
• Its very difficult to get feedback on ideas one
  person at a time.

5
Purpose

• Publishing is a time-tested way of putting your
  ideas out there for the general audience to
  critique en masse.

6
Purpose

• Moreover, academic publishing is supposed to be
  relatively impartial in judgment.
• Not driven.
• Academic publishing, unlike other media like WWW,
  has some quality controls.
• Any idiot with a server can put totally false
  information up on the Web
• Whats worse is that some people believe it.

7
Purpose

• Over the course of time, academic publishing
  allows the best ideas to remain and solidify into
  the facts that we use today.
• Its an old way but still the best way.
• Incidentally, industry also tries to determine
  which of its ideas are good by sending out for
  review as well.
• These a called patents.

8
Purpose

• I maintain that if you are a part of the academic
  community as a grad student, post doc,
  researcher, or faculty, you have a duty and an
  obligation to participate in the exchange of
  ideas through publishing.
• Otherwise you are just freeloading off the
  ideas generated by others.
• For this society to work, everyones got to pull
  their own weight.

9
Process

• Two different avenues of academic publishing
• (a) Archival journals and (b) conference
  proceedings
• These are both valuable methods of disseminating
  ideas.
• They serve different purposes though.
• Should not be confused with each other.

10
Process

• Conference Papers
• Aimed at rapid dissemination of new ideas.
• Testing the waters by getting quick feedback from
  your peers in a discussion forum.
• Should let you know whether or not youre on the
  right track with an idea.
• Gauge the interest level after your presentation.
• Find out whos else is working on it, whether
  youve missed anything, what else should be
  considered.

11
Process

• Journal Papers
• Lasting archival documents
• Much more thorough than conference papers
• Should have a complete idea fleshed out and
  documented.
• Partial results shouldnt be presented here
• This is something that should be read by others
  50 years from now.

12
Process

• Timelines
• Most large established conferences have a lead
  time of approximately 1 year between the Call for
  Papers and the actual conference.
• Lead time may seem long but is necessary.
• Having had to organize conferences, I can assure
  you the 1 year lead isnt excessive.

13
Process

• Papers need to be collected.
• Abstracts, Invited sessions, Full Papers
• Papers need to be reviewed to screen for quality
  and appropriateness.
• Wouldnt want a blatant ad for company X to show
  upat least for some conferences.
• This review usually involves sending manuscripts
  to researchers in the community and asking for
  their opinion.
• Takes time

14
Process

• Accepted papers need to be put into an
  appropriate program.
• May not want papers on Dishwashers in same
  session as Space Telescopes.
• Takes time
• Accepted papers need to be sent to the
  publishers.
• Preliminary program and proceedings need to be
  created.
• Takes time.

15
Process

• Draft of proceedings has to be reviewed.
• Takes time
• Final proceedings have to be printed or burned on
  CD
• Takes time
• Information has to be distributed to the
  conference registrants

16
Process

• Journal paper timelines are different.
• There is no set time the paper has to appear in
  print.
• The key here is not speed, its quality.
• The paper gets submitted to the Editor of the
  journal.
• That editor then logs the paper in and identifies
  an associate editor
• Suffice it to say that all the following steps
  take time.

17
Process

• Associate editor identifies qualified reviewers
  and manuscript is sent out.
• Associate editor pesters and pesters and pesters
  (and.) the reviewers to try and get the reviews
  on the manuscript back in a timely manner.
• In fairness to the reviewers theyre not getting
  compensated in any way for their work.
• This is the longest step in the process.

18
Process

• Once collecting the reviews, the associate editor
  can either (a) reject the paper, (b) accept the
  paper, or (c) send it back to the author to make
  corrections based on the reviews.
• If (c) then repeat the review process until the
  AE decides (a) or (b).
• Usually not more than one additional round of
  iteration.

19
Process

• The whole process from submission to appearing in
  print takes a long time.
• For example ASME JDSMC has an average time of
  about 2 years.
• About 20 of their papers take over 3 years from
  submission to print.
• Other journals have even longer time scales some
  have shorter ones.
• Depends a lot on editorial staff and quality of
  journal.
• Lesser journals have lower standards and can
  turnaround quicker also fewer submissions.

20
Process

• Therefore, authors have to continue to put ideas
  in the pipeline so that theres stuff coming out.
• However, whats a year or two when youre trying
  to ensure that you publish ideas thatll last for
  several decades?
• The goal is to do things properly rather than
  quickly.

21
Process

• The time for journals is longer than for
  conferences because the quality expectations are
  A LOT higher.
• Personally, my journal papers are usually an
  amalgamation/distillation of ideas that have been
  presented at several conferences.
• That way only the appropriate ideas are presented
  to the journal reviewers and in the appropriate
  manner.

22
Process

• Evidence of the difference in quality level
  between conferences and journals is the
  acceptance rate.
• On average a conference acceptance rate (ACC,
  IMECE) is probably about 80.
• Varies with conference.
• Some CS conferences have acceptance rates closer
  to Journal levels.
• Journal acceptance rates can be 33.
• Used to be for JDSMC. Now theyre around 40-45
• IEEE TAC or SCL is down around 35 as well.

23
Process

• These acceptance rates are part of the reason
  its so important to have a very high quality
  initial submission to a journal.
• The reviewers will take any excuse to reject the
  paper.
• I just handled a paper where the reviewer
  rejected it for not citing stuff properly.
• The burden of proof is on the author to explain,
  not the reviewers to understand.
• Key misconception by young authors who expect the
  reader to figure out what theyre trying to say.

24
Pitfalls

• There are several common ones.
• One of the key ones is shortsightedness or myopia
  of the authors.
• Author is unable to put their work into a broader
  context.
• Author understands what they did but has no idea
  why it is valuable outside of their office or lab.

25
Pitfalls

• By not understanding where the work fits in
  within a broader context, the author has no
  ability to state whether or not this is a
  contribution to the field.
• Even if they do state it is a contribution,
  their ability to back that up is really weak.
• One way to avoid this is to become very familiar
  with the field.
• Thats why literature review is so important it
  allows you to say how/why your work is relevant.

26
Pitfalls

• A key pitfall is the author writing the paper as
  if they were writing to themselves.
• Good authors can put themselves in the
  perspective of a reviewer, or a potential reader
  5-10 years down the line.
• Then they are able to discern whats really
  important.
• What key fact(s) am I trying to leave the reader
  with?
• In what format would s/he like to receive the
  information e.g. paper organization and layout.

27
Pitfalls

• Example
• Students often want to write about exactly what
  they did.in the order they did it.
• Why? Usually because thats the way they were
  taught to do it in lab courses.
• Usually, some of the ancillary details associated
  with the work may not be that relevant in 10
  years.

28
Pitfalls

• E.g. controller implementation details acronyms
  and descriptions.
• Software changes over time algorithms dont.
• Moreover, the chronological order in which
  results were obtained is not necessarily the best
  order in which to recount them.
• Actually, it usually isnt you often find out
  stuff later which can reframe the problem more
  clearly.

29
Pitfalls

• Being able to put oneself in the shoes of the
  reviewer/reader is an invaluable skill in writing
  high quality papers.
• Assuming that the actual work done was high
  quality.
• You cant dress up junk too often.
• I try to imagine that Im a researcher in Europe
  10 years from now who wasnt involved with the
  project and wonder what information would I want
  out of this work.
• Or China, or India, or wherever.
• Main point is that theyve never seen the
  setup/software.

30
Pitfalls

• Another major pitfall, related to the previous
  one, is the lack of perspective of the author.
• They dont know what is the most important part
  of their work.
• All they write about is what they did which may
  not have been what they should have done.
• What are the key questions that should be asked,
  and are those answered here?

31
Pitfalls

• Example
• Currently handling a paper where the author is
  doing engine idle speed control (a well worked
  problem) using throttle actuation.
• Author did not use spark timing in their
  formulation even though its what everyone doing
  idle speed control includes. (i.e. ABV spark).
• The author just did throttle because it was the
  easiest to implement experimentally.
• However, this is of little value because the
  field, including most car companies, has already
  acknowledged that you can do better with 2
  inputs.
• Author lacked perspective

32
Pitfalls

• Final one Ill mention here is lack of clarity.
• There are others but time is short.
• Authors dont pick a central theme and stick to
  it.
• Papers become a patchwork of fragmented ideas and
  results without a central goal.
• Exposition without motivation.

33
Pitfalls

• OKhow do we avoid some of these pitfalls?
• Experience is an excellent teacher.
• Not that you should write a lot of crappy papers
  to figure out how to write good ones.
• Read, read, read, read.
• The more information you can gather about what
  goes on in a field, the better your perspective
  will be on that field.
• Hopefully you can also get an idea of what good
  papers look like

34
Pitfalls

• Listen, listen, listen, listen.
• Attending seminars will also provide perspective
  by understanding whos doing what.
• Same thing for conferences listen to whos doing
  what.
• Participate in the process.
• Journals only show you good papers.
• The best way to figure out what crappy papers
  look like is to review and read a few of them.
• Once you get frustrated a couple times, youll
  make sure someone else wont be thinking _at_!
  about you.

35
Pitfalls

• Participating in the review process will make you
  a reader.
• This makes it easier to write for the reader.
• Youll see the difference between clear vs.
  unclear and myopic vs. experienced
  thinking/writing.
• Think about that kid in Germany, China, or India
  10 years from now.

36
Perks

• There are a lot of advantages to publishing.
• A key one is that you get external feedback on
  whether your idea is a good one or not.
• Remember, generation and dissemination of ideas
  are the fuel that drives the engine of academia

37
Perks

• Editors, associate editors, and reviewers will
  spend many hours on your manuscript.
• Trust me on this one!
• Its basically a free evaluation of your ideas.
• I sure as dont get paid for it!
• Think about how much youd have to pay a
  consulting firm to tell you whether your idea was
  good.

38
Perks

• Acceptance of your ideas should give you personal
  satisfaction.
• Also, you receive recognition as having
  accomplished something that is accepted by the
  broader community.
• This is helpful when youre being evaluated for
  future employment.
• True for industry, govt. labs, or academia.
• They now know you can come up with good ideas and
  sell them.

39
Perks

• One of the most immediate perks is travel to fun
  and exciting places to present your work.

ACC 02 in Alaska
40
Perks

• ARG papers and research have appeared, or will
  appear, at conferences all over the world.
• All over the U.S.
• Europe NL, Germany, Spain, Portugal
• Asia Vietnam, China, Japan
• Not yet Africa, South America or Australia
• ...not yet! ?

41
Perks

• Final perk Ill mention here is that by
  publishing and presenting, you become part of a
  worldwide community.
• Ive got friends that I can look up in nearly any
  city in the U.S.
• There are people all over the world that I can
  stay with, go out to dinner, go out to bars, etc.

42
Conclusions

• Academic publishing is important but very
  rewarding.
• There are good ways and bad ways to try it.
• Actually, the bad ways will never see
  publication.
• Ive given you some things to avoid as well as
  ways to avoid them.
• Please do try to avoid them because you dont get
  a second chance to make a first impression on a
  reviewer. S/he may be a future employer,
  colleague, etc.
• Now go be productive ?