Role of background knowledge in interpreting visualizations?
Visualization to support exploratory analysis?
Fostering visual literacy to support adoption and sustained use?
Can it be solved with current resources?
There are many open issues to do with practices - some of which have promising work going on, other s require new work
A double-edged sword - constraints assist beginners but are a block to innovation
Could be a way in for newcomers or communities with very specific requirements (a few Googles for specific tasks)
Fostering visual literacy to support adoption and sustained use?
Can we coordinate UK viz activity better?
Can we coordinate UK viz activity better?
What common principles/overlaps are there?
Visualization is a collaborative process inherently - the science and the vis expertise
Top
Role of background knowledge in interpreting visualizations?
Role of background knowledge in interpreting visualizations?
People bring domain knowledge to vis. How can we exploit this in designing visualizations?
e.g. statisticians have to understand chemistry conventions and priorities
eg. cholera epidemic example - to interpret it you had to understand meaning of X to signify waterpoints
You want to hide stuff that everyone knows often - redundant (but of course a hidden assmption may be important
What issues does this raise in collaborative contexts (differnt users and backgrounds)?
How well do tools support the capture of multiple interpretations?
This connects to the metadta context issue
Activity around the vis. tool - what is its interface/ interoperability with these? (e.g. linking to the logbook from a data point - reading and writing to it from the viz tool)
Top
Visualization to support exploratory analysis?
Visualization to support (eg. statistical) analysis?
Is analysis a separate step from Vis? Shouldn' they be part of the same integrated envrnt?
User interaction is key - iteration between data, analysis, views. What if?... tests - What is thi s data point?
Challenge is to integrate these elegantly in the UI
eLab book - pervasive capture of all media annotations - want to recover these from the viz. (Soton)
Activity around the vis. tool - what is its interface/ interoperability with these? (e.g. linking to the logbook from a data point - reading and writing to it from the viz tool)
How to capture 'context' of activity around the vis.? - what metadata?
Comb-e-chem ontology for experiments is an example
"Is this helpful?" - the critical process of vis expert and domain expert consultation
Multiple experiments - tracing a series of studies - vis. software is a resource for group memory
Top
How can users and vis experts work effectively together?
Collaborative modelling/consultation
Are there HCI design/evaluation methods particularly well suited to eSci-Vis design?
Could they be 'packaged' in an attractive form for eSci-Vis teams to use?
How well do collaborative vis tools support this at a distance? Or asynchronously?
COVISA scenarios explored this a bit
Requirements change when you start building something tangible to critique - how to support this?
back to visual literacy - the more literate you are, the better able to specify your requirements y ou are
Fostering visual literacy to support adoption and sustained use?
exploiting common representations in a domain is an obvious point of common ground
a way towards the 'Google' idea
Problem with the template/Google idea is that predefined options close off possibility of discovery via new visualizations
If a commnuity works with a small set of visualizations, then a templates approach is fine (a highly developed literacy)
Need to get the balance between working with what people know, and opening their eyes to new possib ilities
stuednt training establishes common ground
Role of background knowledge in interpreting visualizations?
Top
Fostering visual literacy to support adoption and sustained use?
Fostering visual literacy to support adoption and sustained use?
What are the obstacles to scientists using visualization tools?
What are the 'visual literacy' skills associated with eSci-Vis?
For a given problem, knowing what 'templates' are powerful tools to use (or customise) (template = data, renderings, ... parameters
How easy to share such a configuration across tools?
ontologies/ interoperability standards could help
More complex when you're using specific HW
Vis is more complex than slides! How to make an intuitive 'walk up and use' (Google) interface?
Vis is a huge area - but domains have their own specialised tools -- encod more background knowledge
One could define 'template' designs for a given community - 'shrink-wrapping' for wider consumption ('turn-key solutions')
'Training wheels' in HCI literature
Gurus - Tinkers - Gardeners spectrum of power user (HCI lit)
Statistics community - recognise the need to understand an area's priorities and restrict the offer ings to them
e.g. IRIS Explorer web user interface delivers a service - web form => VRML rendering (local control over the result)
Intuitive interfaces for computational steering?
Again, will have to be a restricted UI that makes assumptions - a steppnig stone that may draw peop el in deeper
IV conference talk with exmples of scientists' drawings in history - Pat Hanrahan (Stanford) - eg trees with annotation and distortion to emphasise points
other examples: graphs, tables, Role of artistry
Viz lies session on poor examples
always easier to critique - harder to recommend (cf. voluminous UI guidelines that are hard to con textualise)
Keller and Keller - book on good visual cues
Top
Can we coordinate UK viz activity better?
Can we coordinate UK viz activity better?
Vis. brings in many different people - hard to focus events
Many scientists think of vis as a presentation tool only (+ think they know it), not an exploratory analytic tool
Some scientists will attend vis. specific events
Others (most?) might only attend disciplinary events, and then attend a vis. tutorial etc (which could provide hands-on experience etc)
Computer-support staff events another obvious target
Vis. is a tool that crosses disciplines - makes it hard to share lessons
Top