Ivan Stojmenovic site.uottawa/~ivan

1. How to write research articlesin computing and related engineering disciplinesIEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, VOL. 21, NO. 2, FEBRUARY 2010, 1-3 Ivan Stojmenovic www.site.uottawa.ca/~ivan

2. Motivation to write this article • Make research presentations as easy to comprehend as absolutely possible • Vast majority of papers does NOT follow basic (and hopefully logical) structure • The same structure is valid for thesis work, conference and journal publications, or technical reports for research sponsors. • Prosense project • Article and direct impact by Veljko Milutinovic • Experience: Consequences (duplication) of a bad presentation of my most cited contribution

3. Basic structure • the problem statement, • existing solutions, • new solution(s), • assumptions and limitations, • analysis, simulation and • comparison with best competing solutions.

4. Motivation = publication • motivation for writing article is its publication. • Acceptance factors: quality, politics, and luck. • This article helps to improve the quality of an article by improving its presentation. • Luck: selection of reviewers, their bias or fairness, acceptance rate etc. • Politics: Many papers are published even in top conferences and journals because of influential authors, despite possible poor presentation or even poor quality.

5. Politics is important • The probability of acceptance depends on factors such as the scope of journal/conf., networking or personal links to editors and committee members, and the quality of paper. • the quality is an input for political decisions. • Political decisions are often made by reviewers, based on visibility and reputation of authors, competition with other articles etc. • a single low quality article opens the doors for criticism and negative recommendations, even for an upcoming high quality article

6. Excellent presentation and risk? • A novel idea, well presented, can be claimed to be too simple or even trivial by reviewers. • simplicity of any original idea is its advantage • Reviewers may not admit they do not understand • well-understood idea may be identified as already existing, is easier to identify the drawbacks and criticize by reviewers • If problem statement has a new name, it is difficult to identify proper referees for the article and/or to properly judge its contribution.

7. Four key parts of an article • the title, abstract, introduction, main body • Describe contribution 4 times, with increasing word counts: 10, 100, 1000, 10000. • Each of them should be self-contained and complete to the greatest possible extent. • Reader attention: 80% title only, 15% abstract, 4% introduction, 1% full text.. • Clear and appealing text in one part increases the chances that a reader will go to the next part, and eventually use and cite the work. Or not go further and give positive review… • Readers: thesis examiners, reviewers of conferences and journal submissions. They all have limited time to spend on a particular article and time should be used wisely.

8. Research = Problem solving ? • Development ? • Implementation ? • Example:‘software architecture’ problem ? • Example: teaching recursion in first year CS • (problem: what is the best teaching method?) • Example: Gabriel graph application • (problem: what existing structure will be the best for routing with guaranteed delivery?) • Example: survey article: existing surveys? New taxonomy?

9. Grant applications • "If we knew what it was we were doing, it would not be called research, would it?" (Einstein). • Objectives ? • Milestones ? • Dates to finish research and meet objectives? • Are grants given for development, or for research ?

10. Science vs Engineering • Science: new solution aims to be the best for given problem under given assumptions and scenarios, and for a given model • Literature review then needs to be thorough to identify proper competing solutions • Innovation/Engineering: new solution should be good solution that works for given problem in a practical setting

11. Innovation: Performance evaluation • Validation more important than comparison • Less emphasis to originality and competitors • Modeling and assumptions in ‘practical’ problem ? • Validation by simulation may be questionable • Best solution does not mean a valid solution in practice - only under given model and assumptions • I. Stojmenovic, Simulation in sensor networks…, IEEE Communications Magazine, December 2008.

12. Selection of ‘Title’ • The selected title should • enable the expert to figure out the essence of the basic idea(s) and the main contribution(s), even without reading the paper [M1]. • induce the reader to think deeply over the "philosophy" of the contribution [M1]. • [M1] = Veljko Milutinovic • Help Google Scholar search

13. Abstract: Protecting the Contribution • the most important part of a research article. • New idea should be protected in the abstract to the greatest possible extent. • A clear abstract is the key to having the work properly credited in other people’s work. • Other literature reviews could ‘cut and paste’ • Should contain: problem statement, existing solutions, essence of new solution with advantages, analysis done, comparison with existing solutions

14. Content of Introduction or Chapter 1 • General overview • Precise problem statement, good intuition level • Existing solutions and their criticism • Contributions (proposed solutions; why they are expected to be better; essence of the idea(s) used in proposed solutions); • Conditions, assumptions and limitations • Analysis (theoretical, experimental, simulations, implementations,…); under what conditions and scenarios is the new solution the best one? • contribution of the thesis author to the thesis work, and the publications/submissions from the thesis • The structure and content of the rest of the document

15. Literature review:The next section/chapter • Full literature review (except well-known facts) • thorough literature review on the problem • (unless it has hundreds solutions..) • What are competing solutions? • Give a reason to dismiss a solution, or compare, analytically or by simulation (critical literature review) • clear separation line, between existing work and new ideas • contribution could be significant, but short in text.

16. Literature review • Describe briefly the main ideas of competing solutions, • especially for those similar to new solution (what is the difference?) • Explain how an algorithm (existing or new) works before analyzing or criticizing it • Good and thorough literature review means reduced contribution (a risk)

17. Time for literature review ? • It takes long time to get first result in a new field • Subsequent papers take much shorter time • Team work helps • New problem statement means small literature review, and no competing solutions • New environment for a problem statement does not mean that existing solutions for other environments are not applicable • Exiting solutions are often ‘painted’ in literature for new environments

18. The remaining chapters/sections • should present new contributions (including conditions, assumptions, and limitations, where appropriate) and their analysis. • same items should be presented in full, • preferably in the same order

19. Figures, examples • a figure may be worth a thousands words • Important new concepts, new ideas, should be illustrated by examples and figures • Examples not trivial, but meaningful and helpful. • Figures, examples, diagrams… should not be overly repetitive. • Reading only the figure captions of the paper should almost substitute for the first rough reading of the entire paper • parameter values and protocol names must be clearly visible and/or listed in the caption

20. Overview for long solutions • If solution is long and detailed then • Summarize it first while giving main ideas at intuitive level • Possibly with an example, • Then present all steps in detail.

21. More advices • search for scenarios where new solution is the best • A smaller but justified claim is better than a large unfounded claim • include all the possible criticisms of your own idea and contribution directly in the article • show full control of the problem, solutions, and their performance

22. Conclusions and references • Some people read only the abstract and the conclusion. • The most important part of the conclusion section is to list future work that can be done using the results of the current article. • follow a +-+ pattern in the introduction and the main text (positive, negative, positive). • References like [BMSU] for Bose, Morin, S.., U..

23. Overall flow and appearance • smooth transition from topic to topic • Within each of the abstract, introduction, or main text, repetitions should be avoided. • pay more attention to the language used and the overall appearance. • extract a positive impression for the subjective part of the overall evaluation is by showing the overall care taken

24. Related work on presentation • Veljko Milutinovic: slides • most deal primarily with language, grammar and formatting issues • Alba: Introduction, Problems (this part includes literature review), Resolution methods, Experiments, Results, Conclusions, and References. • Woodford: begins from asking whether the time is right for writing, to analyzing and possibly answering the examiner’s remarks

25. Other fields • Chemistry, Psychology, …: • Abstract, Introduction, Methods (or Theoretical Methods), Results and Discussion, and Conclusions. • Mathematics: less rigorous • papers are collections of theorems and proofs, and every known proof from other sources is cited in the text where needed • Cheney: ‘Mathematics is preeminent in its striving for • absolute precision in its formal written text. • use thecorrect word at the proper place and carefully construct each sentence. • against the use of slang, colloquialisms, and other non-standard linguistic devices’. • Applicable to CS & E.

26. Cheney: ‘Use English descriptions and English text in preference to mathematical symbolism wherever possible.’ unit disk graph G over set of sensors S: G=(S, V), V={(u, v), |uv| ≤ R}). two sensors can communicate with each other if and only if the distance between them is at most R, where R is the common transmission radius. mathematical symbolism is by its nature INTIMIDATING, even to mathematicians. the impact of possible misprints ?? Violate with good reason More on math

27. ‘Systems’ papers • Discuss reality, lessons learned, choices made • Implemented? Used? Practical importance? • Otherwise, do the ideas justify publication now? • What author and reader should learn ? • How generally applicable are lessons ? • Realistic assumptions? • Alternatives at various points ? • Formal model? Support by deep theorem?

28. Conclusion • Writing articles in a form that other people can understand is a very slow process • You will be rewarded by a broader distribution and a greater understanding of your ideas within the community