Reading Papers• Need to read efficiently

– A survey paper or grant could have 100-200 references.

– Read 50+ articles a year • Need to know how your papers is going to be

read. (???)

Reading Papers

• What do you want to learn?– Just generally interested in the topic– Want to know if the subject is relevant to your

research– Want to know what the authors’ concluded– Want to learn about a new area of research

• How are you going to find out efficiently?

Reading Order• Title – think about what it means• Authors – do you know them?• Keywords – semi-useful (ACM categories) • Abstract – read more?• Conclusions or Introduction – read more?• Introduction or Conclusions –read more?• Results• Methods or background• Background or methods

Reading Order

• Is the topic relevant? (title and abstract)• Is it new? (abstract)• What did they find? (abstract and conclusions)• What does it mean? (abstract and conclusions)• Are the results relevant? (conclusions and results

(and methods))• Are the results trustworthy? (results and methods)• Should I use their techniques? (methods)

Reading Papers

• Write on the paper– Important points– Weaknesses– Counter-arguments– Relevant papers– Promising citations

General Writing - Structure

• Broad and general to narrow and specific– In the whole paper– In sections– In paragraphs

• The read should know what’s coming next– Use topic sentences

• Introductory Sections• Introductory Paragraphs

General Writing - Flow• The reader should usually know what’s coming

next – the paper should flow– Sentence to sentence– Paragraph to paragraph– Section to section

• Flow– Topics – the reader should know when and why you

are changing topics– Ideas – the reader should know when and why you

are introducing a new idea

Title and Abstract

• Get the reader interested– Present the subject– Present the significance– Present the work– Present the results

Title

• Try to include:– Research subject (often implied)– Problem– Work– Results

• Can’t always get them all in.• Avoid cute

Abstract

• Write it Last or First?– First – because if you can’t you don’t know what

you are doing– Last – because the emphasis may change as you

write the paper• Do both! Write a rough version first, the write

a new version last.

Abstract (short form)

• Four to six sentences:– Problem/significance/context– Your work– Results/impact

• Adjust to your venue– Often longer– Extended abstracts are something different.

Abstract - exampleCurrent fault attacks against public key cryptography focus on traditional schemes, such as RSA and ECC, and, to a lesser extent, on primitives such as XTR. However, bilinear maps, or pairings, have presented theorists with a new and increasingly popular way of constructing cryptographic protocols. Most notably, this has resulted in efficient methods for Identity Based Encryption (IBE). Since identity-based cryptography seems an ideal partner for identity aware devices such as smart-cards, in this paper, we examine the security of concrete pairing instantiations in terms of fault attack.Computers, IEEE Transactions onPublication Date: Sept. 2006Volume: 55,  Issue: 9On page(s): 1075- 1080

Abstract - exampleIn the mouse trigeminal pathway, sensory inputs from distinct facial structures, such as whiskers or lower jaw and lip, are topographically mapped onto the somatosensory cortex through relay stations in the thalamus and hindbrain. In the developing hindbrain, the mechanisms generating such maps remain elusive. We found that in the principal sensory nucleus, the whisker-related map is contributed by rhombomere 3–derived neurons, whereas the rhombomere 2–derived progeny supply the lower jaw and lip representation. Moreover, early Hoxa2 expression in

neuroepithelium prevents the trigeminal nerve from ectopically projecting to the cerebellum, whereas late expression in the principal sensory nucleus promotes selective arborization of whisker-related afferents and topographic connectivity to the thalamus. Hoxa2 inactivation further results in the absence of whisker-related maps in the postnatal brain. Thus, Hoxa2- and rhombomere 3–dependent cues determine the whisker area map and are required for the assembly of the whisker-to-barrel somatosensory circuit.

Science 8 September 2006:Vol. 313. no. 5792, pp. 1408 - 1413

Abstract• Work and results should be precise, refutable

– Avoid• “We study …”• “We examine …”• “We tried X and it works.” Works?

– Use• “We show …”• “We implemented X under the following conditions … The

results show that X has 10% fewer errors…”

Abstract - Examples“…in this paper, we examine the security of concrete pairing instantiations in terms of fault attack.” (end of abstract) – And the results are?

“We use simulation to show that attack traffic can only degrade legitimate traffic to a limited extent, …” - Reasonable, but what is limited?

“Thus, Hoxa2- and rhombomere 3–dependent cues determine the whisker area map and are required for the assembly of the whisker-to-barrel somatosensory circuit.”- Determine and required are solid, definite terms

Keywords

• Use ones standard in the field• Use terms you want to introduce

Introduction and Background• Introduction – Get the reader interested

– Why is it important?– What did you do?– What were your results?

• Background– More context– Why is this paper important– NOT a laundry list of related papers– (Move some to the discussion)

Introduction• Why is it important?

– Broad problem/question – -> narrow problem/question– -> (potential) solution– -> result

• Disease X is important. • Disease X has similarities to condition Y. • Condition Y is due to Z.• We hypothesize Z has a similar role in Disease X.• We test…• We find…

Introduction• Other elements to include

– What’s missing from the existing literature (implied or explicit)

“Currently the link between Z and X is unknown”– What your hypothesis/question is (can be explicit)“In this paper we test the hypothesis that…”– What you did (in general terms)“We used method X to …”– The results

Background• Not a laundry list of papers• Only include papers that directly apply to your

research– Strengthen the connections made in the

introduction• Should lead to your research

“Thus, it makes sense to test Z’s role in Disease X.”

Methods

• Replicable– Cite others

• Justify choices– Why did you use Wormstrom’s method and not

Farnsworth’s?• Explain how the experiments test the

hypothesis

Results

1. Present relevant results.2. Present the results that you promised in the

methods section.3. Tell the read what they should see/learn

1. What does the data show – be explicit and exact2. Say it in the text and in the figure captions

4. Figures (w/ captions) and text should both (mostly) stand on their own.

Conclusion/Discussion

• Emphasize the key results• Summarize• Generalizations?• Related problems?• Future work?