Notes on Writing and Style

1. Notes on Writing and Style

2. Styles Verbose or cryptic, flowery or plain, poetic or literal Conventions important � reduce the effort required from readers Disregarding conventions � may distract from the message ( unless that is the message)

3. Science Writing Prosaic Clear, accurate, but not dull Economy � every sentence necessary but not to the point of over condensing Ego less � you are writing for the readers not yourself

4. Scientific Tone Objective and accurate To inform not entertain Do not over qualify � modify every claim with caveats and cautions Never use idioms like �crop up�, �loose track�, �it turned out that�, etc. Use examples if they aid in clarification

5. Scientific Motivation Brief summaries at the beginning and end of each section The connection between one paragraph and the next should be obvious Make sure your reader has sufficient knowledge to understand what follows

6. Other Writing Issues The upper hand � inclusion of offhanded remarks like � �this is a straightforward application �� Write for your dullest readers, as an equal Obfuscation � aim is to give an impression of having done something without actually claiming to have done it Analogies � only worthwhile if it significantly reduces the work of understanding, most of the time bad analogies lead the reader astray

7. Writing Issues Straw men � indefensible hypothesis posed for the sole purpose of being demolished �it can be argued that databases do not require indexes� Also use to contrast a new idea with some impossibly bad alternative, to put the new idea in a favorable light

8. References Up-to-date Relevant (no padding) Original source First order: books and journal articles Second order: conference article Third order: technical report No private communications or forums ( material cannot be accessed or verified) if you must leave as a footnote not in the bibliography Do not cite support for common knowledge

9. References Careful wording is required when you restate other work. Robinson�s theory suggests that fast access is possible, but he did not perform experiments to confirm his results[22]. Much better Robinson�s theory suggests that fast access is possible [22], but as yet there is no experimental confirmation

10. Unsubstantiated Claims Most user prefer the graphical style of interface. to We believe that �. Another possibility would be a disk-based method, but this approach is unlikely to be successful. Another �, but our experience suggests that �

11. Citation Style References should not be anonymous Other work [6] -> Marsden [6] has � In self-references, readers should know that you are using yourself to support your argument not independent authorities Avoid unnecessary discussion of references, Several authors �., we cite �

12. Citation style Ordinal-number style, name-and-date style, superscripted ordinal numbers, and strings. Use anyone, but use one! Entries ordered By appearance of citation alphabetically

13. Quotation Text from another source If short � enclosed in double quotes If long � set aside in an indented block Long quotations, full material, algorithms, figures may require permission from the publisher and from the author of the original Use of quotes for other reasons is not recommended

14. Acknowledgements Anyone who made a contribution Advice, proofreading, technical support, funding resources Don�t list your family, unless they really contributed to the scientific contents

15. Ethics Not present opinions as fact Distort truths Plagiarize Imply that previously published results are original Papers available on the internet � authors put out an informal publication and becomes accepted as a formal. It is expected that the informal version will be removed

16. Scientific Writing

17. Titles Titles should be concise and informative A New Signature File Scheme based on Multiple-Block Descriptor Files for Indexing Very Large Data Bases (better) Signature File Indexes Based on Multiple-Block Descriptor Files An Investigation of the Effectiveness of Extensions to Standard Ranking Techniques for Large Text Collections (better) Extensions to Ranking Techniques for Large Text Collections

18. Opening Paragraphs Begin well Most care with the opening Bad openings This paper concerns In this paper Distinguish description of existing knowledge from the description of the paper�s contribution

19. Paragraphing Discussion of a single topic or issue Long paragraphs can be an indication that the author has not disentangled his thoughts Readers pay attention to the first lines and last Link paragraphs by reuse of key words or phrases

20. Lists Good, but don�t overuse, only for important information A list of trivia can be more attention grabbing than a paragraph of important information

21. Sentences Simple structure, a line or two long Avoid nested structures In the first stage, the backtracking tokenizer with a two-element retry buffer, errors, including illegal adjacencies as well as unrecognized tokens, are stored on a error stack for collation in to a complete report. (better) The first stage is the backtracking tokenizer with a two-element retry buffer. In this stage possible errors include illegal adjacencies as well as unrecognized tokens; when detected, errors are stored on a stack for collation into a complete report

22. Repetition and Parallelism Text that consists of the same form of sentence used again and again is monotonous. Careful with however, moreover, therefore, hence, thus, and, but and then Complementary (antonyms) concepts should be explained as parallels Access is fast, but at the expense of slow update (better) Access is fast but update is slow

23. Direct Statements The following theorem can now be proved (active) We can now prove the following theorem. Artificial use of verbs Tree structures can be utilized for dynamic storage of terms. Terms can be stored in dynamic tree structures (watch � perform, utilize, achieve, conducted, occurred)

24. Direct Statements �we show� In this paper it is shown that The authors show (Here we can help explain to the reader who is making the contribution) Other times we should not be used When we conducted the experiment it showed that our conjecture was correct (correct) The experiment showed that our conjecture was correct

25. Ambiguity Check carefully The compiler did not accept the program because it contained errors. (better) The program did not compile because it contained errors

26. Qualifiers One per sentence ( might, may, perhaps, possible, likely) It is perhaps possible that the algorithm might fail on unusual input. (better) The algorithm might fail on unusual input.

27. Padding The fact that In general In any case Remove these

28. Misused Words Watch for Which, that, the May, might, can may is for personal choice can to indicate capability Less, few less, continuous quantities (space) Fewer, discrete quantities (errors)

29. Misused Words Affect, effect Effect � consequence of an action Affect � influence, as in outcomes Alternate, alternative, choice Alternate � switch between Alternative � something that can be chosen Choice � more than one alternative Note, if there is but one alternative, there is no choice

30. Overuse of Words Same word in the same sentence is annoying. Redundancy Adding together -> adding After the end of -> after In the region of -> approximately

31. Tense Most text past or present Present used for eternal truths The algorithm has complexity � not the algorithm had complexity In references may have to past tense as also in describing work and outcomes � the ideas were tested �.

32. Others Abbreviations - best none Acronyms � use CPU not C.P.U Limit � may confuse reader Sexist language � get rid of pronouns and recast the sentence