How to Write a Good

Conf/Journal Papers
By
Buyung Sofiarto Munir
 01
What is your
target?
International conference
● “International conference” is the place that you
make an oral/poster presentation in front of the
experts, and thus get an interaction with other
researchers much more easily.
● A short paper (2-15 pages) is usually published
in the proceedings, and some good papers are
invited to publish their revised edition in a
journal
● Compared to journals, publication is much
quicker.
International journals
● 4 – 40 pages of publication of
“completed” work
● Must contain an exactly “new” thing with
good evidence that it’s correct/valid.
● There’s no oral communication with
reviewers and potential readers; so your
journal paper must contain everything
that is needed
● Impact factor is important, but avoiding
bogus/fake journals is much more
important!
What is the Impact Factor (IF)?
● The average annual number of citations per article published
● For example, the 2011 impact factor for a journal is calculated as follows:
● A = the number of times articles published in 2009 and 2010 were cited in
indexed journals during 2011
● B = the number of "citable items" (usually articles, reviews, proceedings or
notes; not editorials and letters-to-the-Editor) published in 2009 and 2010
● 2011 impact factor = A/B
● e.g. (600 citations) / (150+150 articles) = 2.000
Try these for finding Impact Factor
● SCImago Journal Rank (SJR)
www.scimagojr.com

● SNIP (Source Normalized Impact per Paper)

www.journalindicators.com
Influences on Impact Factors: Subject Area
Editor & Reviewers are your first reader
● When you submit your paper, the most
important audience/reader are editors
and reviewers.
○ They have the right to accept or reject your paper

○ They are experts in your areas, but don’t think they

are experts of your topic.
● You have to convince them that your
paper is worth accepting.
02
Organization
of research
paper
The importance of story telling
● First, we have to admit that not many of others will
have the same perspective with us. Our research,
however important it looks to our eyes, will be likely
to be boring, wasteful, or unimportant to others, if
they are not persuaded otherwise.
● Human beings live in context. So the authors must
give a good convincing context of your research to
the reader.
● Yes, the author has to persuade the reader /
reviewer that your work is worth reading/publishing.
That’s your effort to make.
Why your writing may not be understood?
● Lack of the necessary definition of
technical terms
● Lack of appropriate description of
backgrounds (social and technical)
that would guarantee the
importance of your work
● Lack of logic
● Simplicity is preferred in research
writing, but it may not work well
in educational writing. Why?
 Do and Don’t in writing
Do
● Use the full verb form instead, e.g. they are, is not,
cannot
● Use language such that the ideas are presented clearly
without any ambiguity. Do not write stories.
● Present your ideas and methods used in the research.
Use to the point observations to justify your research.
● Make sure the data, figures, formula etc you are using is
correct. Check the sources from where you might have
taken the data.
● Try to write in a sequence so that the readers can
understand the flow of the research carried out.
● Ensure you have carried out your tests with sufficient
number of test cases.
● Maintain all the sources and carefully, mention all of
them. Cite the references properly.
● Use plain and simple style of writing. Do not use
“flowery” language. Readers will prefer simple and
understandable language rather than high vocabulary
words.
● Check for nay spelling, punctuation and grammatical
errors.
Dont
● Do not use shortened verb forms (contractions),
such as they're, isn't, can't
● Do not make use of words like “in my opinion” or “I
believe” or “I think”. Instead, use words like, “It is
recommended”.
● Don’t use first person pronouns ("I", "we", "me", "us",
"my", and "our").
● Do not start sentences with “Well,…just don’t”
● Do not make up stories and add unnecessary details
just to increase the length of the paper or to justify
any point.
● Do not make general statements. Present you
findings without giving too many overview from
different papers and sources.
● Do not exceed the mentioned word limit. Follow the
guidelines specified.
● Do not cite Wikipedia as a reference.
● Do not forget to mention the references of the
source for supporting material.
Organization of a paper
● Title of the paper
● Authors, designations, and addresses
● Abstract (100 to 150 words)
● Keywords
● Sections describing the work
○ Introduction (Background)
○ Previous Work (Literature Survey)
○ Proposed Work (Algorithm, Design, Methodology)
○ Analysis (Complexity Analysis, Quantitative Analysis, Statistical Analysis, …)
○ Implementation and Results
○ Conclusions
Abstract
● Abstract is independent of the body of your paper, and must have an
informative information on your paper.
● A good abstract is a miniature of your paper.
Introduction
● Before, reviewers, who are your first readers, go back and forth of
your paper, but now (due to PDF) they start reading from the
beginning to the end.
● If the first impression of your paper is poor, they read your paper to
find defects and flaws. → First impression of your paper comes from
your title, abstract, and introduction
● Smart flow of Introduction
○ 1st paragraph: Societal backgrounds and issues
○ 2nd paragraph: Academic/technical backgrounds
○ 3rd paragraph: Technical problem/difficulties
○ 4th paragraph: Introducing your issue in this paper and how to solve it
○ 5th paragraph: the summary of your result and impact/contribution
○ 6th paragraph: the organization of the paper
Literature Survey (Review)

● Brief description of the existing body of work

● Citations to published work
○ In [2], Rosetti and Longfellow described the meaning of life.
Wordsworth presented a different view point in [3]
● Bring out the specific advantage of your work
w.r.t. published work.
○ However, the Rosetti-Longfellow formula [2] fails for
technologies below 0.2 micron. In this paper, we extend their
formula to the deep submicron domain.
Sections and Subsections

● Organize each section into subsections and

(possibly) sub-subsections
1. Introduction
1.1 Problem Description
1.1.1 Combinatorial Explosion
1.1.2 Dynamic Programming
1.2 Organization of the paper
In Section 2, we summarize the previous work in this area. In
Section 3, we present a new algorithm for …
Main body of work

● Organize your work into sections and sub-

sections.
● For example,
3. New Algorithm for Graph Partitioning
3.1 Genetic Algorithm
3.2 Data Structures
3.2 Crossover Operator
3.3 Mutation Operator
3.4 Convergence Criterion
Experiments: Quantitative Analysis

● Quantitative Analysis
○ Present numerical results: memory consumption, time
consumption, size of the chip, clock rating, power dissipation, …
○ Present tables: Run-time of your algorithm for several
benchmark examples
○ Present graphs: Data size vs. Time Complexity, Chip Area vs.
Clock Frequency
○ Present improvement figures: Our proposed algorithm can
improve time complexity up to 30-40% reduction; our
optimization algorithm resulted in a 20% reduction in chip area
for the sp292 benchmark circuit …
Experiments: Comparative Analysis

● Compare two different approaches to the same

problem
○ Tabulate results for two different heuristics
● Compare the performance for two different
parameters
○ Tabulate results for two different settings, two different datasets
and so on.
● Graphic Items should be as clear as possible.
○ Figures include Pictures, Photographs, Graphs, Plots …
About Figures/Tables
● Number all the figures and tables
● Provide captions for all figures and tables
○ Figure 3. The proposed algorithm
● Refer to each figure and table.
○ In Figure 3, we show the proposed algorithm for fast finding of association rules
based on the scoring function in [4].
● Make a clear illustration
● Define symbols/abbreviations in legends
● Clarify label axes
● Use same units as in text
● Photo authenticity or obtain permission to use previously published
material
About the ‘Conclusions’
● State what was achieved in the paper
● If possible, we should state some definite conclusions
○ We presented two algorithms, A1 and A2, for the association rule mining
problem. Our experimental results indicate that A1 outperforms A2 in
terms of rule quality, but requires about 80% more time than A2 in most
cases.
● We should be straightforward about the limitations of
your work
● Based on the limitation, we should point out directions for
further work
References
Be sure that the format is correct
[1] T. Theeramunkong, Applying passage in web text mining., Int. J. Intell. Syst. 19
(1-2) (2004) 149–158.
[2] K. Sriphaew, T. Theeramunkong, Revealing topic-based relationship among
documents using association rule mining., in Proceedings of the International
Conference on Artificial Intelligence and Applications, IASTED/ACTA Press,
2005, pp. 112–117.
[3] R. Agrawal, H. Mannila, R. Srikant, H. Toivonen, A. I. Verkamo, Fast discovery of
association rules., in Proceedings of Advances in Knowledge Discovery and
Data Mining, AAAI/MIT Press, 1996, pp. 307–328.
 Research ethics – Against FFP
● Fabrication
○ Fabrication is the intentional act of making up data or results and recording
or reporting them.
● Falsification
○ Falsification is manipulating research materials, equipment, or processes, or
changing or omitting/suppressing data or results without scientific or
statistical justification, such that the research is not accurately represented
in the research record. This would include the "misrepresentation of
uncertainty" during statistical analysis of the data.
● Plagiarism
○ Plagiarism is 'the act of representing as one's own original work the creative
works of another, without appropriate acknowledgment of the author or
source'. The site provides practical advice on when and how to properly
acknowledge other people's ideas and work.
○ The term 'creative works' includes: published and unpublished written
documents, interpretations, computer software, designs, music, sounds,
images, photographs, and ideas or ideological frameworks gained through
working with another person or in a group. These works may be in print
and/or electronic media.
Concluding
● When you start your research, try to write a draft of a research paper:
○ You will find that you cannot write many parts in it. That’s OK. The parts that you cannot
write is exactly what you have to do in your research.
● In particular, try to write “introduction.” It is convincing? You can ask your
supervisor, your friends, etc. to read it and ask for their comments.
● To know the current state-of-the art,
○ To use Google and download as many papers as possible (at least more than 100), and skim
them as quickly as possible.
○ To rely on “review articles”
■ ACM Review Transactions
● If you do these, but still you think you cannot find the current state-of-the-
art, it will mean that you are making too sharp focus: broaden your eye.
Thanks Do you have any questions?
buyungsm@gmail.com

CREDITS:
This presentation template was created by
Slidesgo, including icons by Flaticon, infographics
& images by Freepik