ECE 992 Ubicomp: Paper review guidelines

1
ECE 992 UbicompPaper review guidelines

• Andrew Kun

2
Outline

• Research steps
• Paper review guidelines

3
Research steps

• State problem(s)
• State goal(s)
• Propose hypotheses
• Propose steps to test hypotheses
• Explain how problem(s), goal(s) and hypotheses
  fit into existing knowledge
• Produce results of testing hypotheses
• Explain results
• Evaluate research
• State new problems

4
Problem statement

• Question Why?
• Grounded in existing knowledge
• Scientific vs. engineering problems
• New knowledge vs. new application of existing
  knowledge

5
Examples of problems

• Scientific
• How does the accuracy of a speech recognizer of
  an in-car application affect driving
  performance?
• Engineering
• The response of RFID tags is not described in
  sufficient detail for deployment in many
  ubiquitous computing applications.

6
Is this really a problem, or a?

• we needed to deploy a system that
• We did not have the capability to implement the
  algorithm in C.
• Our goal was to

7
Grounded in existing knowledge

• New knowledge builds on existing
• Knowledge generated by you will produce new
  problems
• Sources books, dissertations, theses,
  peer-reviewed articles, technical reports
  (reputation?), maybe websites (reputation?)
• Must stay current knowledge can become obsolete

8
Grounded in existing knowledge
9
Grounded in existing knowledge

• What if you do not ground in existing knowledge
• The relationship of the number of storks in a
  particular area to the number of newborn babies
  in that area is not well understood.
• However storks do not bring babies!

10
Goal statement

• Question What?
• Solve problem?
• There is no model of human-computer spoken
  interaction for multi-threaded dialogues.

11
Hypothesis

• Tentative proposed explanation of facts,
  observations
• Testable
• I dont have a hypothesis, I just explore.

12
Propose steps to test hypotheses

• Question How?
• Approach

13
Explain

• Explain how problem(s), goal(s) and hypotheses
  fit into existing knowledge
• Background
• New? Possible? Storks vs. babies?

14
Produce results

• Results of your approach

15
Explain results

• Your major contribution

16
Evaluate research

• Conclusion
• Reached goals? Tested hypotheses?

17
State new problems

• Future work?
• Scientific research doesnt have an end.

18
Guns, Germs and Steel

• Can you identify RSs in Diamonds book?

19
Outline

• Research steps
• Paper review guidelines

20
You are the review panel

• Lead reviewer
• Scribe
• Reviewers

21
Guidelines lead reviewer

• Preparation
• Presentation touching on RSs
• Surprise us!
• Discussion
• Presentation initiates discussion

22
Guidelines lead reviewer

• RSs 1-4
• Problem(s), goal(s), hypotheses, approach
• RS 5
• Background
• RS 6, 7
• Results and explanation
• RS 8
• Evaluation
• RS 9
• Future steps

23
Guidelines reviewers

• Preparation
• Powerpoint slide(s) touching on RSs
• Discussion - your take on
• Paper
• Lead reviewers presentation
• RS 9 Future steps
• Relation to your work/interests
• Help scribe create review

24
Guidelines reviewers

• RSs 1-4
• Problem(s), goal(s), hypotheses, approach
• RS 5
• Background
• RS 6, 7
• Results and explanation
• RS 8
• Evaluation
• RS 9
• Future steps

25
Guidelines scribe

• Slides
• Strengths, weaknesses, suggestions, relevant
  progress since paper was written
• Review notes at the end of class

26
Discussion ground rules

• Youre expected to interrupt
• Be polite but honest
• Reviews loaded on laptop use them

27
Deliverables

• Presentations (PowerPoint)
• Lead reviewer long (before class)
• Reviewers short (before class)
• Scribe overview of discussion (end of class)

28
Logistics

• Need 2 laptops lead scribe
• Projector mine OK