1. A Bird’s Eye on Academic Writing
科技论文写作
A Bird’s Eye on Academic Writing

2. A Few Words Different disciplines have different writing standards.
Social Science is more demanding than Engineering.
Writing is a life-long learning experience.
Don’t expect to leave the room as a writing expert.
It is enjoyable and rewarding.
Our pragmatic goal is to:
Give you a clear sense of academic writing.
Help you avoid common mistakes in writing.
Pave a road for you to keep improving your skills.
© Dr. Bo Yuan /08/2010

3. Research
Research
The search for knowledge
A systematic investigation to establish facts
Motivated by natural observation or personal interests.
Formulate questions and develop hypotheses.
Design experiments to gather evidence to make a claim.
Justify their hypothesis to peers and community.
How?
© Dr. Bo Yuan /08/2010

4. Writing Research is cumulative. Personal Writing Academic Writing
Built on accepted assumptions and previous work done.
Scrutinized by members of the scientific community.
Personal Writing
Diary, Blog, Letters …
Descriptive, Narrative
Academic Writing
To construct knowledge claims and test ideas.
Analytical, Argumentative
© Dr. Bo Yuan /08/2010

5. Why publication? “Publish or Perish” Contribute to knowledge
Pressure in academia
The primary way to judge a researcher
H-index
Contribute to knowledge
Tell other people (not yourself) an exciting story.
Personal satisfaction
Get credits and citations.
© Dr. Bo Yuan /08/2010 5

6. Types of Papers Research vs. Review Theoretical vs. Experimental
Methods vs. Applications
Long vs. Short
Published vs. Unpublished
© Dr. Bo Yuan /08/2010

7. Where to publish? Journal Conference Books Workshop
© Dr. Bo Yuan /08/2010 7

8. Targeting a Conference
Impact vs. Acceptance Rate
Tricky Choice
May have separate rates for long/short papers.
Extras:
Special Sessions, Workshops, Tutorials
Time & Location
Can you meet the deadline?
Will you be available for the conference?
Is the conference venue attractive?
© Dr. Bo Yuan /08/2010 8

9. Targeting a Journal Prepare a short list of candidates.
Reputable Publishers
Recommendation by Colleagues
Read the instructions for authors.
Aim & Scope
Paper Type & Length
Frequency of Publication
Skim and scan the latest issues.
Popular Topics, Quality, Conventions
Reviewing Period
© Dr. Bo Yuan /08/2010

10. Myths & Facts Myth  Writing is a frustrating task.
Fact  It could be a lot of fun once you can manage it.
Myth  Every researcher knows how to write.
Fact  Good writing skills need years of hard training.
Myth  There are secrete recipes.
Fact  Yes, but they are just guidelines.
© Dr. Bo Yuan /08/2010

11. Plagiarism!!! A serious offence Self-plagiarism
Lack of quotation or citation
Worse than copyright infringement.
Self-plagiarism
Reuse of (nearly) identical contents
Multiple submission
Don’t take any chances!
It will be revealed sooner or later.
Ruin your reputation and credit.
© Dr. Bo Yuan /08/2010 11

12. Other Misconducts Multiple Submission
Identical Work in Different Languages
Data Manipulation
Insufficient New Contents
Unjustified Authorships
© Dr. Bo Yuan /08/2010

13. Presentation Matters Many papers are badly written.
Good ideas get buried …
Papers are to be read by other people.
Clarity, Clarity, Clarity!
Show your seriousness and preciseness.
© Dr. Bo Yuan /08/2010 13

14. Word Processors MS Word LaTex
WYSIWYG (easy to learn)
Third Party Support: EndNote, MathType
Expensive
LaTex
A markup language (takes considerable time to learn)
Better at handling math equations & large files.
Free
Many publishers provide templates in both formats.
In my opinion, it is just a personal preference.
© Dr. Bo Yuan /08/2010

15. Title Capture the eyeballs of potential readers. Bad Titles
Too short →Not specific enough
Too long →Difficult to follow
Good Titles
Informative
Key Words
Match the content
© Dr. Bo Yuan /08/2010 15

16. Title An informative title highlights: A Good Title A Bad Title
The topic of the study
The nature of the study
The scope of the study
A Good Title
A Comparison of Two- and Three-point Belt Restraint Systems in Frontal Crashes
A Bad Title
Belt Restraint Systems in Frontal Crashes
© Dr. Bo Yuan /08/2010

17. Authorship Responsibility vs. Credit Order of Authorship
Make significant intellectual contribution.
Participate in writing and approve the final version.
Order of Authorship
Simple but could be troublesome.
Alphabetical Order
Not all contributions justify authorship.
© Dr. Bo Yuan /08/2010 17

18. Abstract The most challenging part of a paper!
Sell your work in 50 to 150 words.
Many people will only read your abstract!
General Structure
What is the research question?
Why is the scope of the study?
What is your idea?
What is your achievement?
© Dr. Bo Yuan /08/2010 18

19. Abstract Its focus is not on what will be found in the paper.
Use key words and function words.
Avoid empty phrases and sentences:
In this study …
Something is a very interesting area of study.
Avoid non-informative verbs:
Something is described /discussed / presented.
© Dr. Bo Yuan /08/2010

20. Introduction Step 1a: Introduce the background.
Reviewers may not be experts in your domain.
Step 1b: Related work (literature review)
Present existing research work in a systematic way.
Don’t simply put on a list of papers.
Step 2: The missing part (the gap)
Your research question!
Step 3: Preview of the paper
© Dr. Bo Yuan /08/2010 20

21. Introduction Step 1a: Step 1b: increasing interest, growing concern
great importance, much effort
Step 1b:
Past tense for the reference to single studies
Someone et al. (2000) carried out an extensive analysis …
Present Perfect for reference to areas of study
A lot of effort has been devoted to … (Someone et al., 2005)
Present Tense for reference to general knowledge
Something is an effective technique for … (Someone et al., 1990)
© Dr. Bo Yuan /08/2010

22. Introduction Step 2: Step 3:
However, the exact mechanism remains unclear.
Although … , there is still a lack of …
Step 3:
The objective of this study is to …
In this paper, we show …
The structure of this paper is organized as follows.
In the next section, we briefly review …
Experiment results are presented in section V.
Find a few papers and try to identify these steps.
© Dr. Bo Yuan /08/2010

23. Methodology State the motivation clearly.
Highlight the principle and mechanism.
Give a formal definition/specification.
Show an intuitive example.
Use tables and graphs whenever possible.
© Dr. Bo Yuan /08/2010 23

24. Experiments This is where you put your idea under the test.
Present the details of your experiment rigorously.
Materials, Data, Algorithms
Pre- and Post-Processing
Control Parameters, Random Factors
Reviewers will keep an eye on:
Is your experiment design valid?
Is your experiment sufficient to support your claim?
Is your experiment reproducible?
© Dr. Bo Yuan /08/2010

25. Results State the findings of study. Define the scope and limitation.
Use factual statements with necessary explanations.
Past Tense
Define the scope and limitation.
Don’t exaggerate!
Use visualization tools for large amount of data.
Error bars, Box plots …
Matlab can do a good job for you.
Some analysis of the data would be a plus.
Can significantly increase the research value of your work.
© Dr. Bo Yuan /08/2010

26. Discussion Interpret your results.
Explain your own research and indicate its limitations.
Relate your results to previous work.
however, by contrast …
Explain discrepancies in your results.
There may be something unexpected.
Raise new questions and make speculations.
seem, suggest, difficult, probably …
© Dr. Bo Yuan /08/2010

27. Conclusion Final chance to impress the reviewers!
Not a simple repeat of what you have done.
It usually contains:
Motivation & Purpose (research question)
Major findings and achievements
Extensions (don’t throw out really brilliant ideas!)
Potential contributions to related areas
Make the strongest claims possible.
© Dr. Bo Yuan /08/2010 27

28. Acknowledgement Financial Support Technical Support Helpful People
Grants, Scholarships …
Technical Support
Source Code
Dataset
Facilitates
Helpful People
Anonymous Reviewers
Colleagues
© Dr. Bo Yuan /08/2010 28

29. Reference Every claim (not yours) should be referred.
Every reference should be cited in the paper.
Do not cite resources that you did not read.
Style: IEEE, ACM, LNCS …
Bibliography Management: Endnote
© Dr. Bo Yuan /08/2010 29

30. Lists Table of Contents List of Figures List of Tables List of Symbols
Appendix
Pseudo-Code, Proof, Tables, Figures
© Dr. Bo Yuan /08/2010

31. Language Present vs. Past Active vs. Passive It/This/Which … We/Our …
Long vs. Short
Typos
© Dr. Bo Yuan /08/2010 31

32. Self Critique Don’t pretend to be flawless. For example:
There are always refutations and arguments.
For example:
What are the pros and cons of your approach?
Why didn’t you do it differently?
Have you built a strong case for your methodology?
May write one or two sentences for each question.
This is to answer anticipated critiques from reviewers!
Be smart on how much to write and where to fit.
© Dr. Bo Yuan /08/2010

33. Coherence & Fluency It is enjoyable to read a well written paper.
Like reading a fascinating story.
Bad papers are fragmented and obscure.
Make people lost and frustrated.
The key is to make the paper:
Consistent & Indicative
Progressive & Linked
Tell readers what they expect and look forward to!
© Dr. Bo Yuan /08/2010

34. Good Practice A forecasting paragraph at the beginning of your paper
Link words at the start of paragraphs
A conclusive sentence at the end of each paragraph
Clear and unambiguous key terms throughout the paper
An exact match between your aim and your claim
A few clear sentences marking key steps or turning points
© Dr. Bo Yuan /08/2010

35. Checklist Title and Abstract Spelling and Grammar References
Terms and Variables
Tables, Figures, Equations
Format and Page Limit
© Dr. Bo Yuan /08/2010

36. The Reviewing Process Timeline Reviewing Outcomes
Usually 1~3 months but could be much longer.
Reviewing
3~5 reviewers
Single-blind vs. Double-blind
Outcomes
Journal
Accept, Minor Revision, Major Revision, Reject
Conference
Accept, Reject
© Dr. Bo Yuan /08/2010

37. Got Comments Back … A very tough time!
Nobody likes critiques.
Constructive vs. Destructive vs. Hostile
Go through the comments word for word.
Be Positive
Reviewers are trying to help improve your paper.
Mm, I should have explained this point better …
Be Polite
Reviewers have devoted their time.
You may also be a reviewer one day.
© Dr. Bo Yuan /08/2010 37

38. Why are papers rejected?
Your study did not examine an issue considered important.
Your study was not original.
Your research design was problematic.
The conclusions drawn from your data were not justified.
There was a lack of details and replicability.
Your paper was badly written.
© Dr. Bo Yuan /08/2010

39. Reply to the Editor
You may be given a chance to response to the comments.
Thank the editor/reviewers for the useful feedback.
Say that you have done the revision thoroughly.
Address each comment separately.
What you have added?
What you have better explained?
Why no change has been made?
Discuss it with a senior colleague before submission.
© Dr. Bo Yuan /08/2010

40. Conference Trip Registration Visa Application Flight Accommodation
$ 300 – 500
Visa Application
Flight
Accommodation
$ 50 – 100 per room per night
Local Transport & Amenity
Maps, Timetable, Changes
© Dr. Bo Yuan /08/2010

41. Oral vs. Poster Depends on conferences. Oral Presentation Poster
Limited time & audience
One-way communication
Can be very impressive!
Poster
Focus on interaction.
More potential audience
© Dr. Bo Yuan /08/2010 41

42. Slides Flashy vs. Succinct Use light and soft background.
Leave enough space between lines.
Minimum 16 points font
Pictures, Video Clips, Animations …
Slides are not your lecture notes!
© Dr. Bo Yuan /08/2010

43. How to give a talk? Don’t panic! Speak loudly at normal speech rate.
Keep an eye on your audience.
Use voice and pause smartly to address key points.
Manage your time well.
Practice publicly in advance.
© Dr. Bo Yuan /08/2010 43

44. How to answer questions?
If it is a simple one …
Let them finish their questions!
Be humble and respectful.
If it is a challenging one …
Be honest!
Stick to the question.
Tell them what you know and don’t know.
If you don’t even understand …
Ask them to repeat their questions politely.
“Shall we discuss this issue later?”
© Dr. Bo Yuan /08/2010 44

45. How to ask questions? Think before asking …
No enough time for in-depth discussion.
Ask questions that may be of interest to other audience.
You are not the only one having questions!
Keep your questions short.
One question at a time.
Be polite!
Avoid phrasing that would antagonize the speaker.
Leave controversial ones after the talk.
© Dr. Bo Yuan /08/2010

46. How to present a poster? Don’t simply hang up your papers!
Be concise: 15 seconds rule.
A picture is worth a thousand words.
HCI: Color, Font size, Layout …
Be interactive: introduce your work in 30 seconds!
Don’t leave your poster unattended!
© Dr. Bo Yuan /08/2010 46

47. What to do at conferences?
A good excuse to take a few days off!
Gain experience and confidence.
Communication
Networking
Banquet & Social Events
© Dr. Bo Yuan /08/2010 47

48. Conference Tips Plan Ahead Be Active! Membership Financial Support
Conference Calender
Preview the conference program for interesting talks.
Be Active!
Chat with people and ask about their work.
Don’t be shy about your English!
Membership
It pays to join ACM & IEEE as a student.
Financial Support
Student Travel Grant
© Dr. Bo Yuan /08/2010 48

49. Get Connected Conferences, Workshops, Seminars …
Mailing List & Discussion Groups
Official Web Sites of Research Communities
Personal Web Sites of Big Names
Build your own web site!
My Space:
Google Scholar:
© Dr. Bo Yuan /08/2010 49

50. Grant Research work may cost heaps of money.
Need to justify your idea and get your research funded.
Government, Industry, Private Funds …
Grants  Research  Publications  Grants
Life as an academic staff:
1/4: Writing grant proposals
1/4: Teaching, supervision
1/4: Professional activities and administration
1/4: Research
© Dr. Bo Yuan /08/2010

51. Writing a Proposal
Writing grant proposals is a key skill for survival.
Even a student may need a proposal for scholarships.
Help you sell a wonderful thing that does not exist yet.
Like a business plan (from 10 pages to hundreds of pages).
Worth from 10K to millions of dollars.
General Structure
Background (Literature Review)
Objectives and Research Questions
Work Plan & Expected Outcomes
Track Record
© Dr. Bo Yuan /08/2010

52. Tips for Grant Proposals
Unlike a paper, you don’t have the results yet.
Make it as attractive as possible!
Key Factors:
Is the proposed research problem significant?
Is the idea interesting and innovative?
Is the work plan detailed and practical?
Are you capable of carrying it out?
© Dr. Bo Yuan /08/2010

53. Common Mistakes Sadly, most of the proposals get rejected.
A fact that you should be used to.
Don’t be desperate!
Too Ambitious
Be focused (don’t pretend that you can do everything).
Too Vague
Need to be refined and polished.
Too Complacent
Be confident but humble.
© Dr. Bo Yuan /08/2010

54. It is better to … Always make your idea distinctive.
New Challenge
New Method or Theory
New Application
Always have a plan B.
Research is about exploring the unknown world.
Unpredictability is inevitable.
Always have well defined outcomes.
Show your preciseness and seriousness.
© Dr. Bo Yuan /08/2010

55. Time Management Getting 90% done is only half way through the journey.
Research = 1 (Inspiration) + 0s (Persistence)
The first draft can be finished within days.
Keep revising for weeks and get sick of it.
Some Tips:
Efficiency First: Do it when you have the strongest impulse.
Get yourself refreshed when necessary.
Keep a record of your progress.
Milestone, Checkpoint
© Dr. Bo Yuan /08/2010

56. Writers’ Group To provide a forum for discussion of writing.
To provide a framework for the writing process.
To develop and exchange effective practices.
Giving and receiving feedbacks on writing.
Sharing information about conferences and journals.
Analyzing the feedbacks from reviewers.
© Dr. Bo Yuan /08/2010

57. Much More @ WWW A Collection of Research Tips
How to do Research at the MIT AI Lab
Scientific Writing
Use Google!
© Dr. Bo Yuan /08/2010 57

58. Take Home Message Start early! Highlight your core idea.
Explain your work logically.
Make your contributions crystal clear.
Ask friends to read your paper before submission.
© Dr. Bo Yuan /08/2010 58

59. Practice makes perfect!
Read well-written papers back and forth.
Write down your ideas and thoughts.
Prepare work reports regularly.
Learn from the feedbacks from your advisor.
You will be more confident with your thesis!
© Dr. Bo Yuan /08/2010 59

60. Research is more than a job. It is an obsession.
© Dr. Bo Yuan /08/2010