The Quest for Assets Evelyne Viegas Microsoft Research

  • Slides: 7
Download presentation
The Quest for Assets Evelyne Viegas Microsoft Research evelynev@microsoft. com

The Quest for Assets Evelyne Viegas Microsoft Research evelynev@microsoft. com

Enabling and Advancing Internet Research at Scale Vision – Enable the Next Generation Web,

Enabling and Advancing Internet Research at Scale Vision – Enable the Next Generation Web, working with Academia, stakeholders from industry, government, and internet consumers/innovators, to build the Information Highway with an Intelligent and Ubiquitous Safe Web DATA > INFORMATION > KNOWLEDGE Challenge du Jour – How to make real world large scale data available to researchers to perform valid experimentation, while maintaining the privacy of users? NSF Cyber Trust PI Meeting 2008

Accelerating Search in Academic Research 2006 Accelerating Search in Academic Research • 200 proposals

Accelerating Search in Academic Research 2006 Accelerating Search in Academic Research • 200 proposals from 36 countries o • o o Search RFP Awards o o Search assets (15 million search queries + click through) PII (including inadvertent) removed Provided under a limited data licensing agreement Increased quota to the Search API NSF Cyber Trust PI Meeting 2008

Virtual Earth Academic Research Collaboration 2007 Virtual Earth Academic Research Collaboration • o •

Virtual Earth Academic Research Collaboration 2007 Virtual Earth Academic Research Collaboration • o • o Ground Images Provided under a limited data licensing agreement Virtual Earth Awards NSF Cyber Trust PI Meeting 2008

Beyond Search – Semantic Computing and Internet Economics 2008 • o o Search Summit

Beyond Search – Semantic Computing and Internet Economics 2008 • o o Search Summit 2007 Search RFP 06 projects review The Quest for Assets – the Good the Bad and the Wanted Beyond Search – Semantic Computing and Internet Economics o o Search and ad assets (100 million search queries + click through) PII removed Provided under a limited data licensing agreement Increased quota to the Search API NSF Cyber Trust PI Meeting 2008

Enabling and Advancing Internet Research at Scale • o o Data Confidentiality 2007 NSF

Enabling and Advancing Internet Research at Scale • o o Data Confidentiality 2007 NSF supported, Co-sponsored by IBM, Microsoft, NSF Participation from 13 federal agencies; 7 industries; 18 universities o o o (a) better government: reliable technology preventing agencies from accidentally compromising privacy (b) better science: access to more and richer data through private data analysis (c) better industry: personalised search; vendor analysis/testing/bug reporting Private Lives and Public Policies: Confidentiality and Accessibility of Government Statistics (Committee on National Statistics, NRC and the Social Science Research Council, National Academy Press, 1993) Expanding Access to Research Data: Reconciling Risks and Opportunities (Committee on National Statistics, NRC and the Social Science Research Council, National Academy Press, 2005) NSF Cyber Trust PI Meeting 2008

Safe Interactive Data Access • Enable Academic Internet Research with Large Scale Real World

Safe Interactive Data Access • Enable Academic Internet Research with Large Scale Real World Data: quite a challenge! • In need of a Safe Interactive data access framework’ maximizing scientific research AND user’s privacy o “sanatized” slice of data + licensing agreement on data -> maximizes privacy o privacy-preserving data analysis -> maximizes research AND user’s privacy o privacy-enabling data creation -> focus on data use vs. access NSF Cyber Trust PI Meeting 2008