Writing a Journal or Conference Paper¶
What You Should Know Before Writing a Paper¶
Clarity in academic writing is good; concision is good; logical organization is good. But none of these matter if your writing is not seen by readers to be valuable to them. It's commonly thought that value is a matter not of writing but of content, that making the writing clear and logical will let value take care of itself. In fact, successful academic texts, including successful grant applications, rely on a host of writing techniques specific to value, not just communicating value but also creating it. A widely recommended video on this topic features Dr. Larry McEnerney, who for 30 years directed the University of Chicago's writing program, which specializes in advanced academic and professional writing. Every MItHRIL member is encouraged to dedicate time to watch it, despite its length.
Come Up with a Project¶
Talk to your supervisor(s) about a project topic. Be sure to define the scope of the topic that will be described in the paper: different venues have different expectations on the amount of content that should be in a submitted manuscript; journal papers usually consist of multiple experiments and are longer in length, while conference papers are generally shorter, around 5-6 pages, and only have room for one experiment.
Establish an Order of Authorship¶
The order of authors listed on papers holds significance in academic writing. The first author generally does the most work on the project and has the greatest influence on the direction and contents of the paper. Subsequent authors are ordered by the quantity of contributions to the project (usually a subjective determination). Supervisors are generally listed as anchoring (last) authors. Discuss authorship early on, preferably even before the project commences, to convey workload expectations and contributions and avoid disputes later.
If there is a perceived tie in the quantity/quality of contributions of the two authors who have contributed the most, a common mechanism is a footnote following each author's name stating that each contributed to the work equally, with the order of presentation then decided by a mutually agreed-upon spectator competition noted on the paper (examples on the source page included a Super Smash Bros match, a Disney Princess sing-a-long judged by a neutral party, a rap battle, a game of Twilight Struggle, a 1v1 StarCraft II match, a piano/guitar duel, and a Nerf battle). If you have any questions or conflicts regarding authorship, feel free to discuss them with Prof. Pan.
Choosing a Good, Catchy Title¶
See https://universityaffairs.ca/career-advice/ask-dr-editor/how-to-write-an-effective-title-for-your-next-journal-article/ for guidance on writing an effective title. The title should convey the novelty or significance of your work in a compact format that is interesting to a potential reader.
Find an Appropriate Document Template¶
Different conference and journal venues have different templates that they would like authors to use. Find and download the appropriate template, and note the rules associated with each.
IEEE Manuscript Template: the IEEE manuscript template will usually be used by our group, as most of our publications are targeted toward IEEE journals and conferences (https://www.ieee.org/conferences/publishing/templates.html).
| Format | Description |
|---|---|
| Microsoft Word Template | Download from the IEEE templates page above. |
| LaTeX Template | See the IEEEtran HOWTO guide (https://ctan.mirror.rafal.ca/macros/latex/contrib/IEEEtran/IEEEtran_HOWTO.pdf) for the PDF instruction guide, and download the LaTeX and BibTeX templates from the IEEE templates page. |
| Overleaf LaTeX Template | https://www.overleaf.com/gallery/tagged/ieee-official |
Layout the Story¶
Having a narrative for a paper provides scaffolding for ideas that can help you effectively present your work and results. Once you have the results of your experiment(s), think about the overarching story you want to tell. The story can be as simple as: we had an idea; to test this idea we conducted an experiment; we thought the results would lead us to conclude X; however, the results actually showed Y; this is why we think the results showed Y and not X; in the end, we contributed this to science and knowledge; for the future, we want to try these other experiments to test more of this. Laying out the story helps you frame the work for the audience and write a paper that flows better from section to section.
Layout the Sections¶
Once you've laid out the story, frame it in terms of the generally used sections in a paper: Introduction, Background, Contributions, Methods and Procedures, Results, Discussion, Conclusions.
As the publications generated at MItHRIL are technical documents, follow standard conventions for reporting experimental results; see the Technical Writing guide.
Writing¶
Introduction¶
The introduction should establish the context and value of your work: why is the work you are doing important, why is it relevant, and in what context is it important? It should also describe, in a nutshell, what your paper seeks to achieve and by what means.
Literature Review / Background / Prior Work Section¶
"If I have seen further, it is by standing on the shoulders of giants." — Sir Isaac Newton
The literature review highlights prior work that your work directly builds upon. This is not a section to take lightly; don't throw in a bunch of semi-related references thinking that will be fine. A literature review methodically explains the background for your work, sets the scene, helps explain why your work is important and what others have done in the field, and, most crucially, helps bolster your ideas. Craft this section to highlight not only what has been done but what has not been investigated and why this "oversight," which you'll be exploring, is important to examine. Use the literature review to sell your ideas.
Best practices for selecting references:
- Where possible, select more recent citations over older ones, though older references useful for establishing the context of the field are fine.
- Avoid referencing papers from predatory journals and venues; check Beall's list, avoid MDPI journals, and stick with publishers you know (check with your supervisor if in doubt).
- You do not need to go into detail about what each work did; group similar citations together (e.g., "some work has examined the role of haptics in virtual reality [22-24]"), but do go into some detail for citations that are highly relevant.
- Compare your proposed work against prior work, and state what you'll be doing differently.
Contributions¶
The contributions can be their own section or rolled into the introduction. Clearly state the contributions of your work to the state of the art; your contribution could be a method, system, dataset, or set of observations. Describe in a few short statements exactly what you have done over prior work.
Methods and Procedures, Results, Discussion, Conclusions¶
These sections follow standard conventions for the venue and are left for the author to develop for each paper.
References¶
References should be cited in the appropriate format as required by the publication venue.
Write the Abstract Last¶
The abstract should be the last thing you complete when writing a paper. It is a short summary of your work's importance, contributions, results, and discussion, so readers can get a high-level overview before digging into the contents. Check word/character limits imposed by the venue and keep within them.
Submit¶
Submission should never be left until the last minute; things like copyright releases, keywords, forms, and author information all take time to complete, and submission portals have been known to become swamped with uploads and crash near deadlines. Aim to submit a day or more in advance, or a few hours before the deadline at the latest.
Figures¶
Figures require significant thought and effort to adequately convey experimental details; some schools of thought hold that figures should be generated first, with the paper constructed around them. See Figures in Publications, Posters, Reports and Presentations for the lab's full guidance on figure preparation, including the outline method for photos, tools (PowerPoint, Inkscape, GIMP), and colour palette selection.
Over the Page Limit in a LaTeX Manuscript? Read Here.¶
A couple of tricks for fitting manuscripts within page limits when just a tad over:
- Shrink the whitespace after the author block and before the content blocks for IEEE papers:
\IEEEaftertitletext{\vspace{-1\baselineskip}} - Adjust interline spacing by setting a scale value smaller than one (a value of 0.9 is fine; do not shrink too much, or lines will be too crowded):
\renewcommand{\baselinestretch}{scale}
Migration note: the source page's Figures section substantially overlapped with the standalone Figures in Publications, Posters, Reports and Presentations page; the duplicated content has since been trimmed here in favour of a pointer to that page, which is now the canonical source on figures.
