Difference between revisions of "Paper/Thesis Checklist"

Latest revision as of 00:01, 10 November 2015

Here is a list of things to check on your paper drafts before I review them. If you have reasons to differ, ask me!

1. Spell check. You can use ispell or aspell on LaTeX documents.

2. Edit your bibliography:

• Use consistent abbreviations of conferences/journals in conference version to keep short. DAC, ICCAD, etc. You can ignore this if it isn't a common conference or journal to the topic.
• Use full name of conference/journal in journal version. "Design Automation Conference (DAC)", "International Conference on Computer Aided Design (ICCAD)", etc.
• Include page numbers, volumes, and issue number.
• Ensure author are all first initial (maybe middle initial) and last name.
• Use "et al." for more than about 3 authors on conference version.
• Put {} around acronyms and units in title to preserve case.
• MOST IMPORTANT IS TO BE CONSISTENT.

3. Each caption should conclude what the reader should infer about the table or figure. Don't say what the figure is, this information is in the title and axes.

4. Use Section~\ref (or Chapter in thesis), Figure~\ref, Table~\ref and Equation~\eqref for all references.

5. Fix any missing references to sections, figures or the bibliography. These show up as ?? in LaTeX generated PDFs and are also given as warnings by LaTeX.

6. Each paragraph should have a comment before it that summarizes the topic sentence in case the paragraph gets off topic. These should be your original outline.

7. The introduction should try to be one page and end with a summary about the remaining sections.

8. Make a bullet list of the contributions of this paper before the paper summary at the end of the introduction. Reviewers don't want to search for what your contribution is, so make it easy for them!

9. Use LaTeX for all numbers: $2.9mW$, $3GHz$, $4ns$, for example. No spaces before units.

10. Define all acronyms on the first use Like This (LT). There should be a space before the opening parenthesis.

11. The abstract should state in a few sentences: what is the problem, why it is hard and/or important, how we solve it, and our main result. No more and no less.

12. All plots should have titles and axis labels (with units).

13. Avoid color plots or at least have unique line styles so it is clear when printed in black and white. Make sure lines and dots are thick enough to see (Make sure all plots and images have a consistent style too -- having differences looks sloppy.)

14. Never use a bibliography reference as a noun. For example, "In [4], the authors..." or "[4] showed that". IEEE style says not to do this. Instead say "Smith et al. showed that... [4]." with a citation.

15. Avoid unnecessary words (more to come): "so as to" -> "so" "analysis of the effect of X" -> "analysis of X" "consider the fact that X" -> "consider X" (see elements of style)

16. Avoid vague words and phrases (more to come):

• huge (use a number or comparison)
• X is important (say why!)
• X is a major problem (say why!)

17. Avoid subjective words (more to come):

• novel (if the first to do something, say so!)

18. Use active verbs (more to come):

20. Don't start sentences with And and Or. Instead, make a compound sentence.

21. Equations should be used in a sentence including punctuation and describe all variables unless previously defined. An example is a line,

y=mx+b,            (5)

where m is the slope and b is the offset. The first letter after the equation shouldn't be capitalized or indented (by not leaving a blank line in LaTeX) since it isn't starting a new sentence.

22. When low power and high speed are used as adjectives, they should be hyphenated: a low-power, high-speed design compared to something that is low power or high speed. They are hyphenated adjectives. (Let me know if I'm wrong on this one!)

23. Only capitalize proper nouns (names, places, etc.), acronyms being defined or the beginning of a sentence.