Upload
exascale-infolab
View
142
Download
0
Embed Size (px)
DESCRIPTION
Conference talk at ISWC 2011 for the award winning outrageous ideas track, October 2011, Bonn, Germany
Citation preview
An Integrated Socio-‐Technical Crowdsourcing Pla8orm for
Accelera;ng Returns in eScience Karl Aberer, Alexey Boyarsky,
Philippe Cudré-‐Maurox, Gianluca Demar-ni, and Oleg Ruchayskiy
Science
Yesterday Today
GiIed Individuals Collabora;ve Effort
OPERA Collabora;on
Scien;st-‐Computer Symbiosis
• A single scien;st has no more the capacity to process all the informa;on – High complexity of systems and workflows – Various fields of exper;se involved
• New discoveries will emerge from community-‐based socio-‐technical systems
Community-‐based Socio-‐technical Systems
• Such pla8orms will be useful – Locally to the scien;st – By extrac;ng knowledge used globally
• They will enable cross-‐pollina;on – All ar;facts need to be interoperable – Higher order logic to combine them
Science
Tomorrow
Collec;ve Intelligence
What do we need?
• Highly-‐expressive machine-‐readable formats – Ontologies of unprecedented quality – Implicit knowledge available in the head of the experts
• Understanding concepts, assump;ons, phenomena, abstrac;ons
• Create a mental map of a research field • Understand analysis methods
A Giant Crowdsourcing Conceptualiza;on Machine
Towards Self-‐Awareness
• A Scien;fic infrastructure – Complex ontological networks – Capture the scien;fic process – Automate rou;ne opera;ons – Share scien;fic ar;facts
• Experts will train the system with their daily ac;vi;es
An “entropy-‐reduc;on” machine
• Relate en;;es • Provide lineage informa;on • Discriminate conflic;ng informa;on • Reason and infer new data
The Web: a Collec;ve Intelligence engine
• Informa;on systems are not instruments • A catalyst for the scien;fic progress • Reason and combine scien;fic ar;facts at very large scale
• Individual scien;st will not be able to fully appreciate models and methods
Scien;fic progress
Time
Now