Try to have some text between major and minor section headings. For example, don't have Section 2 followed immediately by Section 2.1, with no text in between.
Don't end a section with a list; add a concluding remark.
Don't go down to subsection detail in the overview of all sections, e.g. describe Section 2.2 only inside Section 2, not in the introduction of the paper.
Always include the affiliation of each author.
Where possible, merge affiliations; use {aba,cdde}@cs.rmit.edu.au for the email addresses.
Avoid (a)...(b)... lists within sentences - consider using a vertical list.
Introduce new terms and acronyms, and tell the reader where to find more information.
Try to describe the context (related work) earlier rather than later.
Discuss the merits/drawbacks of related work, and contrast your approach. Don't just list ten papers.
If reviewers point out any problems, address them before resubmitting.
A good way to find problems is to carefully read the paper out loud.
In titles, don't capitalise minor words such as "a", "at", "for", "and"...
Writing style
Try to avoid first names. Say "Bloggs and Doe [3] have...", rather than "Charlie Bloggs and John Doe [3] have...".
Don't have stand-alone sentences that are all quoted. Say something like: Bloggs points out that, "jsdjfjasdjfajsjfl".
Ensure quotation marks are paired. In LaTeX, use `` and '' (or ").
Avoid abbreviations such as "i.e.", "etc.", "e.g.", the ampersand "&"...
Avoid parentheses. If it's not a necessary part of your sentence, drop it.
Don't confuse:
its/it's ("it's"="it is")
effect/affect
compliment/complement
practice/practise
advice/advise
device/devise
...
For URLs, use the LaTeX url package for URLs; include the protocol (http://), omit the trailing forward slash and index files (index.html/shtml/php...). If space is not an issue, write URLs on a separate line, particularly in the bibliography. For example, write: note={~\\URL: \url{http://....}}
Don't say "Section X discusses"; you discuss things, not the section!
If you're not near-perfect, always run a spell check, including on your reviews - how can someone take your review seriously if it's full of mistakes? For LaTeX, use ispell/aspell; in Vim, use the vimspell package to spellcheck as you write.
If you're using LaTeX, use style and diction; these are available on goanna in the ~saied/bin/ directory.
If you're using Microsoft Word, make use of the spelling and grammar checking functions.
You can also use latex2rtf to convert a LaTeX file to RTF, or latex2html -split 0 mypaper.tex to convert to a single HTML file, and then load this in MS Word for grammar checking.
Avoid bold font.
Don't write "student", "supervisor", "Dr.", "Masters student" etc. in the personal details.
"few people" means "not many people", while "a few people" means "some people".
Don't say:
"a bit"
"a lot off"
"absolutely"
"all about"
"are going to"
"fortunately"
"get"
"in real life"
"in recent years"
"in today's world"
"nowadays"
"on the other hand"
"on top of"
"these days"
"unfortunately"
"would like to"
...
Avoid the first and second person ("I" and "you"); always use the plural "We".
As a rule of thumb, don't include more than two direct quotes in your paper - it reduces the impact of your work.
Avoid abbreviated verb forms such as "don't", "let's", "we've", "it's" ...
Avoid numbers starting a line; in LaTeX, use the tilde character: there are~109 items shown in Section~\ref{Introduction}, Figure~\ref{StateDiagram}.; in MS Word, use the non-breaking space character.
Avoid starting a paper or section by quoting material; it's best to provide some context first.
Instead of saying "we come to the conclusion that", say "we conclude that".
Don't start a sentence with "Also", "And", "But", "Especially", "So", "Specially", "Whereas", "Yet", "Then", ...
Don't use the apostrophe for plurals such as MBs and 1970s.
Don't use the "/" (slash) character, as in "A is better/faster than B". Reword to drop the slash.
Don't wax lyrical! "Today, technology has made life very easy, computers are fast, people are happy...".
"et al." needs a full stop after the "al.".
When something is countable, use "fewer than" rather than "less than".
For numbers less than hundred, try to write the number out in words, especially numbers less than twenty; for example, write "two items" rather than "2 items".
If you're quoting, keep the citation outside the quote, e.g. Bloggs says, "blah blah" [3].
Don't include a space between a footnote and the previous word, e.g. ...Google\footnote{http://google.com}..
Make large numbers easier to read by inserting a small space after every three digits, for example: 850 123 456. In LaTeX, use the \, character. Don't use commas for this - the can be confused with "real" commas.
Put a space after every full stop and comma; there should be no space before such punctuation marks.
Put a space before opening parentheses.
Reword to avoid awkward phrases such as "using the nodes' string pointers".
Say "Related work" rather than "Literature survey".
"setup" is a noun; "set up" is a verb.
Substantiate every assertion; for example, if you say, "Many researchers believe that...", you should provide some supporting references.
There is no need to use "(R)" or "(TM)" in every sentence. You can simply add "All trademarks are the property of their respective owners" in the acknowledgments.
When starting a sentence with "Therefore", "However", "Moreover", and "Hence", follow this word with a comma. Often, you can drop this word without affecting the meaning of your sentence (don't overuse these words).
When talking about the Internet or the Web, always capitalise in this way.
Make sure the text flows logically in one direction. Don't keep the reader switching between multiple concepts.
You can try to address this by thinking carefully about the
structure of the section or chapter, then sticking to this
structure.
A comma or semicolon is a signal to pause; read your sentences out loud and see if these pauses make sense.
A semicolon should have a complete sentence on each side
Thesis structure: In general, a thesis will have the following chapters:
Introduction
Background
Related work (optional)
Contribution 1
Contribution 2
Contribution 3
Contribution 4 (optional)
Conclusions
Try not to vary too much from this layout, as it is what examiners
expect.
If you need to keep referring to something in the appendix, it probably shouldn't go in the appendix.
At the end of each chapter, provide a summary that reminds the reader what problem you address in the chapter; how you addressed it; what the outcomes were; and what will come in the next chapter.
Referencing
Don't be afraid to contact the authors of a paper for further information; cite this as "Joe Bloggs, Personal Communication, 1st January 2003".
Don't cite the author of the LaTeX template you're using!
Cite only what you refer to in the text. If you want a reference to appear, ensure you actually refer to it in the text!
Don't use citations as articles in your sentences; for example, don't say "in [1]" or "[1] contains".
Follow the referencing style specified by the publisher.
When mentioning author names, use surnames only; don't say "John and Mary have shown", nor "John Doe and Mary Contrary have shown", but rather "Doe and Contrary have shown".
In the references, specify all the pages of the paper, not just page you used.
Pay attention to references; they must contain all data, such as conference name, conference location, date, editors, and page number.
Merge adjacent citations: [3,7], not [3][4] or [3],[4]; some publisher templates only support the latter style, in which case there's not much you can do about it!
Citations should appear before the full stop: "[7]. ", not ". [7]".
Keep a space before the citation.
Avoid numbers and citations starting a line; in LaTeX, use the tilde character: blah~\cite{Bloggs:1998}; in MS Word, insert a non-breaking space.
Try to cite reliable sources, such as peer-reviewed journal articles and conference papers. Books that aren't research monographs are less trustworthy. If possible, avoid basing your research on newspaper articles and web pages.
The acknowledgments section of a paper is not a place for thanking:
everyone you know;
your supervisor - they are usually part of the authoring team;
the people who wrote the template; or
the course coordinator!
You can thank a wider range of people in your thesis acknowledgments.
For references using BibTeX, you need to specify preferred capitalisation by enclosing the text in braces. For example, to force "TCP" to be written as all capitals all the time, you need to write "{TCP}".
Citations should be in reference order: [1,7], not [7,1], and [Bloggs, 1994; Zeeshan, 1990];
in practice, this generally means you should use alphabetical
order in your cite{} command
Take care when naming researchers in your text. Correct examples are:
We use the XYZ approach [Bloggs et al., 1994].
We use the XYZ approach of Bloggs et al. [1994].
Use \citep or \citet (or
\cite and \citeasnoun) correctly to achieve this.
When talking about prior work, include citations; when talking
about your own work, ensure this is clear. For example, when
introducing new terminology, ensure it's clear whether you've
invented that term, or whether somebody else did (cite them)
Figures, tables, and equations
Refer to all tables and figures in the text, and explain what they show.
Be consistent in how you refer to figures, tables, sections etc. Personally, I prefer the format "in Figure 3" over "in figure 3", but this is often determined by the guidelines of your publisher. Don't capitalise unnumbered references; for example, say "As discussed in the previous section, we..." (not Section).
Each figure and table should have a caption that tells the reader what it is showing. A caption of one or two words, such as "system block diagram", is not descriptive!
If the figure is not your own, you should clearly note that you have taken it from somewhere else. Ideally, you should try to obtain permission from the copyright holder.
It's generally better to have the graph span the width of the page. With jgraph, specify xaxis size 6cm yaxis size 4cm to control the width and height of the graph. Try to avoid forcing the size from LaTeX - the fonts will look distorted.
In two-column formats, have large figures straddle both columns; in LaTeX, use figure*. In MS Word, insert two section breaks, and set the new middle section to single-column layout.
Place figures and tables at the top of the page, or on a separate page if necessary. In LaTeX, use \begin{figure}[tp].
To place figures and tables in landscape format, use the sidewaysfigure package, and write \begin{sidewaysfigure} in place of \begin{figure}, and \begin{sidewaystable} in place of \begin{table}.
Number all figures, tables, and (particularly in theses) equations.
In theses, use short captions for figures and tables; the entries in the List of Figures and the List of Tables should be typically no longer than one line. To do this in LaTeX, use: \caption[Short caption for LoF or LoT]{Long caption alongside the figure or table \label{fig-Label} }.
In theses, don't be afraid to use more space for figures. Graphs can each be as wide as the full page, and one below the other, or in separate figures.
Where multiple graphs appear in one figures, make sure the individual graphs are labelled to indicate graph "(a)", "(b)", etc. (with jgraph, write this in the graph title). I recommend using separate figures or using left/centre/right or top/middle/bottom in the captions, rather than a/b/c etc.
Be consistent about whether you put captions above or below figures and tables (I prefer below). This may be dictated by the style guidelines of the conference/journal etc.
Text Layout
Avoid widows and orphans, large blanks sections, or repeated blank spaces (in MS Word), repeated full stops etc.
Wide tables and figures can be made to span two columns.
Avoid fancy bullet point styles.
Pay attention to text alignment and justification.
When using LaTeX, ensure you don't have any overfull boxes.
Make sure your figures are clear. A good tool to use is xfig.
Writing content
Always consider what research question(s) your paper or thesis addresses; you may need to adjust your questions or text as the content takes shape.
Misc
If you are asked to review the work of someone you know, tell the conference organisers about the possible conflict of interest.
Don't include your own details when reviewing a paper unless specifically requested.
Avoid plagiarism:
Quote all text that isn't your own - it's not enough to simply cite a paper.
Don't write with your sources in front of you - it makes it difficult to write in your own words.
