A summary. created on 14th May 2018 by sjwang.

[TOC]

Do not use the word unclear.

• Describe the work in an accepted scientific knowledge.

• State the idea that is being investigated.

• Explain what is new about the idea, what is being evaluated, or what contribution the paper is making.

• Justify the theory, by methods such as proof or experiment.

• Use the same style for your name on all your papers.
• Titles of papers and sections should be concise and informative, have specific rather than general terms, and accurately describe the content. Complicated titles with long words are hard to digest.

• length 50-200 words.

• Function: allow readers to judge whether or not the paper is of relevance to them.

• Five-element organization

  1. Background A general statement introducing the broad research area of the particular topic being investigated.
  2. Importance An explanation of the specific problem (difficulty, obstacle, challenge ) to be solved.
  3. Others work A review of existing or standard solutions to this problem and their limitations.
  4. Your work An outline of the proposed new solution.
  5. Your results A summary of how the solution was evaluated and what the outcomes of the evaluation were.

  Five sentences are enough, one for each of the points above.

An expanded version of the abstract, but with a paragraph or two, not a sentence, for each component.

[topic, problem, references to key papers, approach, solution, limitations, outcomes.]

• Motivation:
  1. why the problem is interesting
  2. what the relevant scientific issues are
  3. why the approach taken is a good one, and why the outcomes are significant.
• The reader should understand the scope of the work, and of the problem and the contribution.

Write "Conslusions" not "conclusion".

important results && their significance && limitations

Future work look beyond the current context.

The writing should begin as soon as the research is started.

In a logical structure, may not elegant, can write freely.

Mathematical content, definitions, and the problem statement should be made precise as early as possible.

Introduction—>Body—>Summary—>Introduction—>Abstract

For a novice writer who doesn’t know where to begin, a good starting point is imitation. Choose a paper or thesis whose results are of a similar flavour to your own, analyze its organization, and sketch an organization for your results based on the same pattern. The habit of using similar patterns for papers—their standardization—helps to make them easier to read.

• Economy

  convey information without unnecessary dressing.

  Be egoless—ready to dislike anything you have previously written. Expect to revise several times, perhaps many times.

• Tone. Use we or I.

  Science writing should be objective and accurate.

  • Have one idea per sentence or paragraph and one topic per section.
  • Have a straightforward, logical organization.
  • Use short words.
  • Use short sentences with simple structure.
  • Keep paragraphs short.
  • Avoid buzzwords, clichés, and slang.
  • Avoid excess, in length or style.
  • Omit unnecessary material.
  • Be specific, not vague or abstract.
  • Break these rules if there is a good reason to do so.

• Examples

• Motivation

  Link text together as a narrative—each section should have a clear story to tell. The connection between one paragraph and the next should be obvious. This principle is sometimes expressed as: Tell the reader what you are going to say, then say it, and then tell the reader that you have said it.

  Motivate the reader at each major step in the exposition: explain how a definition (theorem, lemma, whatever) is to be used, or why it is interesting, or how it fits into the overall plan.

  Motivation is essential, but do motivate the right thing. Don’t, say, set the scene by explaining that certain algorithms are too slow for massive databases, and then test your method on a few thousand records; or argue that “Web search needs to be semantically aware” and then propose the use of spelling correction to amend queries. The big-picture topic may well be the inspiration for your overall work, but that does not mean that it is necessarily the right inspiration for a particular paper.

• Voice

  Avoid excessive use of indirect statements (passive voice), particularly descriptions of actions that don’t indicate who or what performs them.

  “achieved”, “carried out”, “conducted”, “done”, “occurred”, and “effected”.

• Straw men

  A straw man is an indefensible hypothesis that an author describes for the sole purpose of criticizing it.

  We believe that most users prefer the graphical style of interface.

  • Avoid nonsense, absurdity, and over-generalization.

    The Web is infinitely large.

  • If there is evidence—a study or proof, not someone else’s opinion—then cite it. Unsubstantiated claims should be clearly noted as such, not dressed up as accepted knowledge.

    Another possibility would be a disk-based method, but our experience suggests that this approach is unlikely to be successful.

• Reference and Citation

  You need to explain the relationship of your new work to existing work, showing how your work builds on previous knowledge and how it differs from contributions in other, relevant papers.

  When discussing a reference with more than two authors, all but the first author’s name can be replaced by “et al.”

The abstract should be written especially well, without an unnecessary word, and the opening sentence should be direct and straightforward.

The first paragraphs must be intelligible to any likely reader.

That is, describe what you have done without the details of how it was done.

Starting an abstract or introduction with “This paper concerns” or “In this paper” often means that results are going to be stated out of context.

• (现状) Most numerical computation is dedicated to manipulation of matrices, but matrix operations are difficult to implement efficiently in current high-level programming languages. (results:) In this paper we describe a new programming lan- guage with matrix manipulation operators.

2. Beginning a paper by stating that a topic is popular or that a problem is important is flat and uninspiring. Such openings succeed in establishing context but fail in motivation.

Ask yourself, whether the opening sentence is clear? easy understand? stating context? assertion? cause and results? positive?

Diversity—in organization, structure, length of sentences and paragraphs, and word choice—helps to keep the reader’s attention.

• A paragraph should discuss a single topic or issue. The outline or the argument is typically captured in the first sentence of each paragraph.
• Every sentence in a paragraph should be on the topic announced in the opening. The last sentence has higher impact than those in the body, so pay attention to sentence order.
• If a long paragraph can be broken, break it.
• Contextual information can be forgotten between paragraphs, and references between paragraphs can be difficult to follow. Do not use THIS/ HIS/… at the beginning of the paragraph.
• Formatted lists is a good choice. (Reserve lists for material that is both significant and in need of enumeration. )

• When using pronouns such as “it”, “this”, and “they”, ensure that the reader knows what is being referred to. Maybe you can use the noun again to specify it.

• Another problem with “it” is that it is overused.

  The machine crashed and it was necessary to reboot it.

• Premature pronouns also lead to difficulties.

  When recursive compilation was first developed, it was impractically slow and required too much memory.

• Sentences should have simple structure, which usually means that they will be no more than a line or two. Don’t say too much all at once.

• Watch out for fractured “if” expressions.

  If <—> when

• Double negatives can be difficult to parse and are ambiguous.

• The opening phrase can, without the context provided by the rest of the sentence, be interpreted as handles for classifying.

• If an “-ing” suffix can be replaced by “-ation of”, as in this example, then it is probably a good idea to do so.

  replaing "-ing" with "-ation of"

Ask yourself what is the Motivation? purpose? necessary? simplified?

• It is better to write “related issues are discussed below” than to write “related issues will be discussed below”.
• Past tense is used for describing work and outcomes. Thus we write “the ideas were tested by experiment”, not “the ideas are tested by experiment”.
• Either past or present tense can be used for discussion of references. Present tense is preferable but past tense can be forced by context.

“however”, “moreover”, “there- fore”, “hence”, “thus”, “and”, “but”, “then”, “so”, “nevertheless”, or “nonetheless”. Likewise, don’t overuse the pattern “First, ... Second, ... Last, ...”.

• Complementary concepts should be explained as parallels, or the reader will have difficulty seeing how the concepts relate to each other. （格式一致， 看起来方便）

  In SIMD, multiple data sets are processed simultaneously by the same instruc- tions, whereas in MIMD multiple data sets are processed simultaneously by different instructions.

• This can be further improved. It is kinder to the reader to move the longer clauses in a list to the end. (并列，长的放后面)

using structure of a sentence, not italiciza or CAPITALS.

###Definitions

The first time they are used.

• Use short words in preference to long, but use an exact long word rather than an approximate short one.
• The “don’t repeat words” rule might apply to creative writing, but not to technical terms that must be clearly understood.
• Don’t make excessive claims about your own work

• Don’t pile qualifiers on top of one another. Within a sentence, use at most one qualifier such as “might”, “may”, “perhaps”, “possibly”, “likely”, “likelihood”, or “could”. Overuse of qualifiers results in text that is lame and timid.

• Qualifiers such as “very” and “quite” should be avoided, because they are in

  effect meaningless.

• Delete —> [Other words of this kind are “totally”, “completely”, “truly”, “highly”, “usually”, “accordingly”, “certainly”, “necessarily”, and “somewhat”. ]

• Which, that, the. Many writers use “which” when “that” is appropriate. Use “which” only when it cannot be replaced by “that”.
• Do not omit that or which.
• Delete “the” is often used unnecessarily;
• Less, fewer. Use “less” for continuous quantities (“it used less space”) and “fewer” for discrete quantities (“there were fewer errors”).
• Affect, effect. The “effect”, or consequence, of an action is to “affect”, or influence, outcomes.
• May, might, can. Many writers use “may” or “might” when they mean “can”. Use “may” to indicate personal choice, and “can” to indicate capability.
• Basic, fundamental. Some writers confuse “basic” with “fundamental”: the former means elementary as well as a foundation. A result should only be described as “basic” if elementary is meant, or readers may get the wrong idea.
• Novel, complex, sophisticated. “Sophisticated” does not mean new or novel, but either advanced or complex. Use “novel” or “complex” if these meanings are intended.
• Continual, continuous. “Continual” is not equivalent to “continuous”. The former means ceaselessly; the latter means unbroken.
• Conversely, inversely, similarly, likewise. Only use “conversely” if the statement that follows really is the opposite of the preceding material. Don’t use “similarly” or “likewise” unless whatever follows has a strong parallel to the preceding material. Some authors use “inversely”, but the meaning is rarely clear; avoid it.
• Fast, quickly, presently, timely, currently. A process is “fast” if it runs quickly; “quickly” means fast, but does not necessarily mean in the near future. Something is “timely” if it is opportune; timeliness has nothing to do with rapidity. Also on the subject of time, “presently” means soon, whereas “currently” means at present.

• Repetition of a word is annoying when it makes the reader feel they have read the same phrase twice, or have read a phrase and an inversion of it.
• Repetition should be eliminated when the same word is used in different senses, or when a word and a synonym of it are used together.

never use 1st, 2nd, etc.

Use they rather than he/she.