Upload
ross-may
View
214
Download
1
Embed Size (px)
Citation preview
Research Seminar – CPIT 695
Writing Skills
1
What is a Paper?
The point of view of the author The idea is to try to give all the information to help others to
judge your contribution; not just the information that leads to judging it in one particular direction or another. – Richard Feynman
The point of view of the reviewer One never notices what has been done; one can only see what
remains to be done. – Marie Curie
The variety of papers
Unpublished draft. Definition: the first shape assumed by a paper. Use: for the author and for his immediate collaborators. Quality control: the author(s).
White paper. Definition: a complete draft on one’s web page or a research proposal. Use: advertising. Quality control: the author(s).
Tech report. Definition: a draft readable by others. Use: either as a time stamp (new draft with a new result), or for the record
(old draft with lots of detail). Quality control: maybe a few colleagues & students.
The variety of papers
Workshop paper. Definition: a record to document a talk. Use: communication among specialists. Quality control: the program committee (if any).
Conference/symposium paper. Definition: a record documenting a talk. Use: communication in a larger community. Quality control: the program committee.
Journal paper. Definition: the author’s final word on a particular topic. Use: archival purposes. Quality control: the journal reviewers.
Also: submitted / revised / final / extended.
The variety of papers
The tech report (revisited) Definition: extended version of a conference / journal paper. Use: typically includes all the proofs (no copyright restrictions here). Quality control: the reviewers.
The draft (revisited) Definition: an unpublished paper. Use: like wine, a draft may improve with age. Quality control: unspecified.
What is a good paper?
Objective what is the goal of this work, what problem is addressed, what
was the current state of the art, who is the work aimed at? Proposal
if this paper presents a new idea, what, in a nutshell, is it? Contributions/claims
what contributions does the paper claim to make? Which one is the most significant?
Evidence Support for claims - Theorems? Case studies? Simulations?
Benchmarks? Does evidence address issues needed to support claims?
What is expected in a short review?
Target: half a page, maximum: one page Clearly-separated (use subheadings) sections covering
Summary (as briefly as you can – two or three sentences) Evidence (what evidence is offered to support the claims?) Strengths (what positive basis is there for publishing/reading
it?) Weaknesses Evaluation (if you were running the conference/journal where
it was published, would you recommend acceptance?) Comments on the quality of the writing
Plus: Queries for discussion
Writing a Paper
Reserve a good portion of time for writing. An hour spent writing is an hour spent considering a problem instead of, say, grappling with a computer. You must spend time away from other distractions to document work and focus your efforts. Like regular exercise, it needs to be done, and you’ll come to enjoy it.
A paper is written for others to read: reviewers, yourself in the future, and people you don’t know yet. (Example: Zhe Yang.)
It should thus reflect all the concerns just mentioned: readability, clarity, etc.
Proceed from what is known to what is (yet) unknown.
Writing Skills
If you put off writing until you've done all the work, you'll lose most of the benefit. Once you start working on a research project, it's a good idea to get into the habit of writing an informal paper explaining what you are up to and what you've learned every day or few days.
The title: The spearhead of your paper. Subject and predicate.
Subject: the topic of your work. Predicate: your spin on this topic.
What and how.
Common Patterns of a Title
Just the facts: noun; adjective noun Qualification: X of Y Enumeration: X and Y; X, Y and Z Empowerment: X {for|with|using|in} Y Action: doing X Substantiation: X: Y X: Linear Logic; The Zipper X of Y: Principles of Programming Languages; The Lord of the
Rings X and Y: Word and Object; Romeo and Juliet; Communication
and Concurrency; There and Back Again X, Y and Z: Call-by-name, call-by-value and the lambda-calculus
Common Patterns of a Title
And more: Partial Evaluation and Automatic Program Generation; Automata, Languages, and Programming
X for Y: Megamodels for O/R/X Mapping; Requirements for Self-adaptation
And more: A Domain-specific Wiki for Programming Technologies; Dynamic Program Analysis for Database Reverse Engineering
X with Y: Compiling with Continuations; Variation Programming with the Choice Calculus
X in Y: Logic in Computer Science; Compilation in the 21st Century
(X and Y) for doing Z: Methods and Tools for Analyzing Software Evolution
Common Patterns of a Title doing X in (Y and Z): Finding Patterns in Biological and Molecular Data doing (X of Y) with Z: Managing the Evolution of F/OSS with Model-Driven
Techniques A in B: C and D: Model Management in the Wild: Foundations and
Applications Recommendations:
Seek linear, one-pass titles. Don’t pile up adjectives and nouns.
Serial Comma Dilemma: is it “X, Y, and Z” (with a last comma) or “X, Y and Z” (with no last comma)? Grammarians are divided. Whatever your choice, be consistent about it.
Your title should be: accurate, simple, punchy, and memorable, and its form should support its content.
The title
It should be informative. It should be concise. It should be catchy / memorable. It needs to be original. It does not need to be funny.
The abstract, minus the last sentence
In this paper an automated bathroom activity monitoring system based on acoustics is described. The system is designed to recognize and classify major activities occurring within a bathroom based on sound. Carefully designed HMM parameters using MFCC features are used for accurate and robust bathroom sound event classification.
“In this paper”, I am stressing my paper and not my contribution.
Experiments to validate the utility of the system were performed firstly in a constrained setting as a proof-of-concept and later in an actual trial involving real people using their bathroom in the normal course of their daily lives. Preliminary results are encouraging with the accuracy rate for most sound categories being above 84 percent.
The last sentence of the abstract
We sincerely believe that the system contributes towards increased understanding of personal hygiene behavioral problems that significantly affect both informal care-giving and clinical care of dementia patients.
This last sentence changes everything!
(And I am no longer speechless.)
The abstract
It should be brief. It should be as informative as possible. It should be updated last (to account for the contents of the
paper). Fact: Many more people will read your abstract than your paper
(e.g., in a bibliographic data base). An abstract is not an ad: there is no need to repeat the name of
your product
Your introduction should
start with a bang; stop with an overview of what follows; mention your prerequisites and notations; and clearly state the achievement of your paper. A compelling example is always good, especially in a
submission. Get your reader’s attention early on and do no waste space with
obvious and general statements.
Start with a bang
“It was a dark and stormy night.” “The house had a slight German accent.” “He awoke—and wanted Mars.” “Whatever your gravity is when you get to the door, remember—
the enemy’s gate is down.” “Mr. and Mrs. Dursley, of number four, Privet Drive, were proud
to say that they were perfectly normal, thank you very much.”
Suggest format of an Introduction
Introduce the area of research Review key publications Identify any gap in the knowledge or questions which
have to be answered Your hypotheses Your aims and objectives, including a brief description
of the methodology How is your research beneficial and to whom
Check for these When drafting the proposal, it is worth asking yourself the following questions
and trying to answer them in the text: Why should anyone spend public, charity or corporate funds on my research
and my research training? Who is my research going to benefit (the stakeholders) or being of use to
(the end users)? Stakeholders and end-users include, for example, the research community, a
professional body or groups of researchers, a particular group of people such as children, older people or doctors, the government, the industry, health services, social workers…... Try and be specific: stating that your research will benefit the world is perhaps a bit too vague!
Is there evidence, for example in the literature, that my research will fill a gap in knowledge or a market demand? How will it build on the existing body of evidence?
Is my research timely, innovative and/or responding to a new trend? How will my research proposal address my training needs as well as, if
applicable, the needs of my current employer?
Research Methodology Research methodology (e.g. techniques, sample size, target populations,
species choice, equipment and data analysis) Explain why is it most appropriate to effectively answer the research question Explaining what alternatives have been considered and why these have been
disregarded You could also point out how your project fits with the research environment
of your prospective place of research and why your chosen university is the best place to conduct your research, in particular if you have access to unique expertise, pieces of equipment or data
Writing Skills Eye Catchers
First words. Last words (of paragraph / section / chapter / thesis). Capitalization (in an abstract).
Pitfalls Exaggerating. Seeking effect for seeking effect: “This paper bridges a much needed gap.” (quoted by Neil Jones) Practicing Cooper’s prose (cf. Mark Twain). Misspelling (always use a spell checker!).
Standing on the shoulders of giants Disparaging earlier work invites the reader to disparage your own work. Better: Think positive!
Writing Skills Think Positive (For example):
The goal is blah, and it would be great to reach it. What has been done so far is remarkable, but it does not quite reach the
goal. In this paper, we take one step further towards the goal.
Development Organized in sections. Should be progressive. Should be as complete as possible. Should be as concise and precise as possible.
Related work Mandatory. Situates the novelty and significance of your work. Where: either part of introduction, or part of conclusion, or stand-alone
section (second or second-to-last).
Writing Skills
What to do? What to write? A section on related work is not a matter of citing. It is a matter of comparing.
Related work Main pitfall: Forgetting or misrepresenting someone else’s
work. In Simon Peyton Jones’s words: Related work should be
celebrated. Bibliographical references
Bibrefs should be used parenthetically, so that they do not interrupt one’s reading.
Writing Skills
The conclusion None in the mathematical tradition. Minimally:
recapitulates the problem and the contribution; assesses the significance of the contribution; suggests / outlines future work.
The pitfalls of one’s future work Often reveals the limits of the authors’ understanding
Writing Skills
References They must be impeccable:
accurate (correct year, etc.); complete (page numbers, etc.). Standard pitfall: misspellings in titles.
Tips for writing a research paper In general, the contents should come first. (But often, spelling things out in the paper tends to clarify its contents.) Approaches:
Top-down: goal and significance first. Bottom-up: results first.
A rule of thumb: put in your paper what you (would) like to find in other’s papers.
As time goes by It sounds silly, but...
remember to date your manuscripts, and remember to update your bibrefs (“This paper is superseded by ...”).
How to write a paper
Do unto others as you would have them do unto you When you have truly exceptional results
P == NP Probably doesn’t matter how you write, people will read
it anyway Most papers are not that exceptional Good writing makes significant difference Better to say little clearly, than saying too much
unclearly
Readability a must
If the paper is not readable, author has not given writing sufficient thought
Two kinds of referees If I cannot understand the paper, it is the writer’s fault If I cannot understand the paper, I cannot reject it
Don’t take chances. Write the paper well.
Badly written papers typically do not get read
Do not irritate the reader
Define notation before use No one is impressed anymore by Greek symbols If you use much notation, make it easy to find
summarize most notation in one place Avoid Using Too Many Acronyms
AUTMA ?! You may know the acronyms well. Do not assume that
the reader does (or cares to)
How to write a theory paper
Unreadability is not the same as formalism
Reader should be able to understand contributions without reading all details
If some proofs are not too important, relegate them to an appendix Proofs are not as worthy as new proof techniques
How to write a systems paper Provide sufficient information to allow people to
reproduce your results people may want to reproduce exciting results do not assume this won’t happen to your paper besides, referees expect the information
Do not provide wrong information Sometimes hard to provide all details in available
space may be forced to omit some information judge what is most essential to the experiments cite a tech report for more information
Discuss related work
Explain how your work relates to state of the art
Discuss relevant past work by other people too
Remember, they may be reviewing your paper. Avoid: The scheme presented by Vaidya performs terribly Prefer: The scheme by Vaidya does not perform as well in
scenario X as it does in scenario Y
Avoid offending people, unless you must
Tell them your shortcomings
If your ideas do not work well in some interesting scenarios, tell the reader
People appreciate a balanced presentation
How to write weak results If results are not that great, come up with better ones Do not hide weak results behind bad writing
Be sure to explain why results are weaker than you expected If you must publish: write well, but may have to go to
second-best conference Only a few conferences in any area are worth publishing in Too many papers in poor conferences bad for your reputation Just because a conference is “IEEE” or “ACM” or
“International” does not mean it is any good If results not good enough for a decent conference,
rethink your problem/solution
Miscellaneous Read some well-written papers
award-winning papers from conferences
Avoid long sentences
If you have nothing to say, say nothing don’t feel obliged to fill up space with useless text if you must fill all available space, use more line
spacing, greater margins, bigger font, bigger figures, anything but drivel
Technical reports
Useful to get early feedback from other researchers
Puts a timestamp on your work
Can include more information / results than might fit in a paper
Application to your thesis work
What are your data? (make a list of answers)
What do you do to them? (make a list of answers)
Which method(s) do you use to do it? (make a list of answers)
And you do all of this for what? (make a list of answers)
Writing Skills: Doing a Revision based on review
Together with the revision Comment the reviews pointwise: the reviewers will appreciate
to see each of their points addressed upfront. Thank the reviewers for their time: they are actually your best
allies. Finally, consider using latexdiff: it is surprisingly useful.
Key Questions to double check your paper
Is the research problem significant/important? NOT: a problem created/imagined by you and no one else cares about it YES: a problem that people care (evidenced by concrete statistics or
examples) Is your research solution significant or addressing technical challenges? (may
be less critical for some type of work) NOT: a solution that is incremental over previous work NOT: a solution that is straightforward/trivial (e.g., simple adoption or
slight adaption of an existing technique is not significant enough, even when you are the first one in doing so)
Is your evaluation justifying the claimed contributions or benefits of your solution? (e.g., faster, detecting more faults, …than existing techniques if any)
Double check by making traceability from your claims listed in your contributions to your research questions to investigate in your evaluation
References CPIT 695 Notes of Dr. Ahmad Barnawi Materials to refer:
Elements of Style Toward Clarity and Grace Mathematical Writing Articles from http://www.michbar.org/generalinfo/plainenglish/
http://cs.au.dk/~danvy/tips-and-tricks/index.html http://web.engr.illinois.edu/~taoxie/publications/writepapers.pdf Learning Latex
Not So Short Introduction of Latex Beamer, Prosper, ACM, etc http://heather.cs.ucdavis.edu/~matloff/latex.html http://www.math.rug.nl/~ernst/latex/ http://www.ams.jhu.edu/~ers/learn-latex/ http://www.tug.org/tutorials/tugindia/ http://www.ats.ucla.edu/stat/latex/
41