CS295J/Week 3.11

tuesday class notes

• discuss some of the most-discussable papers we presented thursday
  • Hua: Measuring effectiveness of graph visualizations: a cognitive load perspective Huang-2009-JIV
  • Caroline: Cognitive Strategies and Eye Movements for Searching Hierarchical Computer Displays
  • Steve: Toward a Perceptual Theory of Flow Visualization
  • Clara: Could I have the Menu Please? An Eye Tracking Study of Design Conventions

• what should we be doing with what we are discussing/learning? why are we doing this?
  • understanding area, understand what qualifies as research contributions, understand the lacey web of human knowledge
    • connection to other things we're reading
    • putting into wiki: very short summary,
  • relationship to proposal
    • hmm, maybe we need to go back and think about revision
    • what do we need to do for that?

• for Thursday: read reviews of proposal, identify some significant criticism, and say how it will be addressed in the revision
  • examples:
    • criticism: literature on XXX is not represented, here are 3 citations
    • response: we have reviewed the literature in this area and added a paragraph to the significance section relating our proposed work to that done in this area
    • criticism: the proposed approach won't work
    • response: we have 1) added a new section summarizing risks of the approach and contingencies if parts of the approach are not successful; 2) we have included preliminary results from 2 new experiments that show that YYY and ZZZ are likely to be feasible;
    • criticism: the intellectual merit of the proposed work is not clear
    • response: we have rewritten portions of the project summary and significance sections to clarify the intellectual merit of the work. Changed portions are in blue.
    • other examples? do one from the actual reviews?
  • let's put on the reviewer's hat to understand some guidelines for writing responses
    • 18 months ago read 10 proposals and wrote up evaluations of them, including yours
    • OR never read yours
    • now has your revised proposal and maybe the reviews from it
    • does NOT have the original version
    • needs to evaluate the revision
    • how?
      • A: find old one, read both in parallel, absorb everything, look through reviews, write up informed evaluation
      • B: go through points in review and check that they have been addressed
    • so write responses to make that as easy for the reviewer as possible
      • a checklist for them to use
      • annotate the review with responses interleaved
      • describe how a change to the proposal addresses the criticism
      • point to where the proposal was changed
      • don't repeat what's in the proposal in the response
      • try to address *every* criticism with a change
      • only argue in the response for the most important battles
      • make it easy for the reviewer to write "The revised proposal completely addresses all the concerns of the reviewers"
      • try not to make other changes that are not essential and that don't directly address reviewer criticisms
      • the exception to this is for new reviewer criticisms that you are reasonably confident you can anticipate

thursday class notes

• dhl notes on papers (where would we cite these in the proposal)
  • small world graph layout: Jenna
    • method for graph layout -- use it in an application
    • how do you know what's important?
  • facial attractiveness: Chen
    • assess user attractiveness and somehow use in interface? (?)
    • create attractive faces for on-screen people or avatars?
  • attention, habituation, conditioning...: Nathan
    • "almost as appeared" :-)
    • title cool -- could we test these model(s)?
    • model(s) may not be computational enough -- principles rather than mathematical
    • no conclusions
    • fig 7 is a nice brain diagram
    • near end they question whether a model at this level can make quantitative predictions
  • On Distinguishing Epistemic from Pragmatic Action: Caroline
    • tetris-rotating paper -- a really hard test case to model, I bet...
    • an interesting addition to a cognitive model
    • users replace one cognitively difficult task with two that are easier
    • "epistemic" -- knoweldge-collecting tasks, not just NOOP's
    • built expert system with cognitive/perceptual/motor components
    • I wonder what order the epistemic discovery happened
  • design of mobile art guide: Clara
    • not sure what to take from this
  • challenge of infovis eval: Diem
    • position paper -- value evaluation
    • what is AVI conf?
  • synchronous interaction among hundreds: Wenjun
    • how to spend one's time is toughest decision...
    • demands of "first life"
    • maybe we can leverage some concepts to make more engaging apps?
    • not clear where to cite this
  • now, where was I? steve
    • great title!
    • interesting idea to help keep track of where users are in a process
    • there is some literature on when to interrupt users least disruptively
    • should we leverage GSR? another kind of data, used as part of interaction
  • data visualization optimization: Hua
    • good followup to Ware paper from last week that proposed the flow part of this work
    • nice contrast between evaluation and theory-based work
    • I wish there were a conclusion
  • BrainFrame: a knowledge vis system for neurosciences: Michael
    • (please link to pdf's, if at all possible)
    • something we could build on for brain researchers to play with
  • knowledge visualizaiton: towards a new discipline and its fields of applications