Standardization Discussion

1
Standardization Discussion

• Jim Myers

2
The Research Process
Its just the Scientific Method
3
The Real Research Process
Non-linear, high-dimensional, coupled,
multi-scale phenomena
Research ?? Application ?? Education
Method
Instrument
Scientific
Laboratory realization
Multiple objectives
4
Cyberenvironments Supporting the Research
Lifecycle
Gap Analysis
Standards / Best practice
Algorithms/ Services
Reference Data
Sensor Data
Publish Share Coordinate Curate Validate Relate
Discover Mine Translate Reference Extract
Provenance
Engineering Views
Annotation
Experiment Setup
Distributed Execution
5
Metadata Protocols APIs

• URIQA SPARQL/SPARQL DAP, (OAI, Z39.50)
• Content Mgmt Systems (resource properties
  model) JSR 170, 283, webDAV
• Handles, LSID, OGSA-Naming, WS-ResourceProperties,
  GFS
• The Semantic Stack doesnt end at SPARQL Rules,
  Logic, and there are open questions
• Are triples sufficient or are 4-5-6- tuples
  needed?
• How can you optimize semantic representations/quer
  ies?
• Are these the same questions as in provenance?

6
Discussion of Provenance Standardization

• Where are we ready to standardize?
• Where are we not ready?
• What are our requirements for a standard?
• Portable across workflow engines?
• Can track single provenance instance across
  engines
• Portable across workflow styles?
• Portable across module technologies?
• Ability to integrate with non-standard provenance
  capabilities?
• Schema, language, API, protocol?