2008-01-09: AQ Cluster Notes

• Introductions

Slides - Stefan Falke

• Overview of AQ Cluster
  • Data Use
  • Information technology
  • Participants in cluster include providers, mediators and users
• Cluster events
  • Hosted workshop on NO2 Data Synthesis - a lot gained from this experience
  • Dataspaces/Datasheets - bring together relavent metadata
  • Eventspace - CA Smoke example
    • Metadata, datasets that are related, web 2.0 content (flickr, blogs, links)harvested by tags

DataSpace

• Explanation by Rudy Husar
• Interoperability stack - need to connect at hardware, software and humanware (web2.0 intro)
• Virtual Workspace Web 2.0 - needs places for resources, tools, forum, space for collaboration on product
  • Eventspace - Application or need (working on one topic), facilitate regulatory process on exceptional events. Eventspace devoted specifically to documenting event.
  • Dataspace - Facilitate metadata linkage to data. Place to record quality, lineage - service orientation allows connection of metadata
• Many needs/many sources of data - need connector pieces (i.e. workspaces)

Workspace Discussion

• John McHenry - first impressions of Web2.0 very positive
  • idea sharing
  • new product output
  • managing project across institutions very effective
• John McHenry - wanted to disucss barriers
• Frank Lindsay - As we go to web 2.0 and we link to services/data that we didn't create the end users need a clear conception to what was being added how it was created. Legal responsibility systems parallel to totally public system
• George Percivall - Good work on Web services, need workspace that presents services to users for decision support
• Steve Young - Need for community of users, customer rating (youtube, amazon) for capturing user feedback
• Danielle Forsyth - Great service what you have and lineage. Policy User - what should I use?
• Rudy Husar - Dataspaces is new concept, policy user need different workspace. Response to Frank Lindsay - appreciated comments on NASA restructuring already looking at data lineage
• Greg Leptok - Barriers. 1) Data lineage what do you extract from process and attach it. Data lineage - chain of organizations processing data and each one needs to add something to the data. Evaluation goes both ways from users and from scientist on how data can be used. 2) Reached limitation of protocols - several aspects of data that don't know how to access. Moderation - not harvesting anything
• Steve Kempler - Scope things that seem to be barriers by prototyping
• , Lots of capabilities and problems that need to be solved. Wants a particular example and then see what the outcome is. Moving this to other applications.
• Stefan - Data Summit defining "Problems" and then looking to see how this can be used for the problem.
• Frank Lindsay - Activities should proceed until aprehended. Can't plan everything, just throw some things out and see if it's useful. People are focused on a data - delightfully surprised how it is used in different ways
• Steve Young - Amazon published lists of favorites by anyone
• Rudy Husar - How lucky are we that we have amazons in perpetual beta. Use human intellect for what it's good for. Amazon Mechanical Turk
• Steve Kempler - Same mechanisms for quality, can also be help files, tools,