Panel 6 - Legal and Social Implications of Shared Collections

Panel Questions:

• For your domain, what are the legal requirements related to data management (Sarbanes-Oxley, IRB, HIPAA, IP)?
• What implications does sharing of data impose on data management systems?
• Can legal requirements be expressed as policies that automate enforcement?
• What are the issues (legal, proprietary, accessibility) behind internationally shared collections?

Ann Zimmerman, University of Michigan [PowerPoint]

opendata.si.umich.edu

• why is it hard to share data ?
• i-school master programs
• paper co-authorship for data providers
• data papers !

Noshir Contractor, Northwestern University [PowerPoint]

slides worth a look - some keywords

• self-interest
• social and resource exchange
• mutual interest and collective action
• theory of contagion
• theory of balance
• homophily
• proximity
• co-evolution

sharing data and knowledge can expose researcher to avalanche of requests for more information - competent jerks vs lovable fools

Melissa Cragin [PowerPoint]

• small science research and data sharing issues - slide well worth looking at - different perspective to big science

Panel 5 - Developing Assessment Criteria

Panel Questions:

• Who defines the required assessment criteria?
• How is a consensus reached on the required assessment criteria?
• Can an international standard form the basis for assessment criteria?
• Are assessment criteria dependent upon the organizing community?

Mark Conrad, NARA LCDRG and ISO MOIMS-rac [PowerPoint]

NARA's ERA - 163 high-level requirements -some beyond current state-of-the-art

NCAST - Nara's technology research group

• expand scope of advanced research
• bring new technologies and methods into NARA
• identify needed expertise and assist NARA to secure it

Who defines required assessment criteria ?

• Acts - Federal, Presidential, FOI
• Communities - president, congress, courts, other fed agencies, public
• archival principles eg arrangement, provenance

Jane Greenberg, Dryad [PowerPoint] (UNC Metadata Research Centre)

• one-stop deposition/access for data objects supporting published research in evolutionary biology
• data package metadata includes (1) article metadata (2) datafile metadata eg genbank object, treebase object
• hierarchy of goals - preservation, discovery, sharing, synthesis
• HIVE - vocabulary services to researchers (Helping Interdisciplinary Vocabulary Engineering) - uses SKOS

John Graybeal, Ocean Observatories Initiative

Steve Richard, Arizona Geological Society [PowerPoint]

Panel 4 - DataNet Federation

Panel Questions:

• How challenges are addressed by your data management infrastructure?
• What are the current practices for data management, curation, preservation?
• What are the interoperability mechanisms for federating registries, ontologies, and processing models (workflows)?
• "What do you see to be the biggest challenges and opportunities with respect to evolution of a DataNet Federation?"
• "What is your vision for a DataNet Federation?"

Bill Michener, DataOne - see pilot catalog for earth observations

http://prototype.dataone.org/pilotcatalog/

• priority - user-friendly tools for researchers eg automated metadata acquisition tools
• more investment needed in documenting researcher requirements for tools

Sayeed Choudhury, Data Conservancy [PowerPoint]

Reagan Moore, DataNet Federation Consortiuim [PowerPoint]

Policy-based data environments:

• Purpose - reason a collection is assembled
• Properties - attributes needed to ensure the purpose
• Policies - control for ensuring maintenance of properties
• Procedures - functions that implement the policies
• State information - results of applying the procedures
• Assessment criteria - validation that state information conforms to the desired purpose
• Federation - controlled sharing of logical name spaces

Goal - Generic Infrastructure

• Manage all stages of the data life cycle
• Data organization
• Data processing pipelines
• Collection creation
• Data sharing
• Data publication
• Data preservation
• Create reference collection against which future information and knowledge is compared
• Each stage uses similar storage, arrangement, description, and access mechanisms

Panel 3 - Large-Scale Data Management Challenges (bottom-up drivers for infrastructure):

Panel Questions:

• What technical solutions exist that meet your academic project requirements?
• What technical solutions exist that meet corporate requirements?
• What requirements are unique to the academic environment?
• What requirements are unique to the corporate environment?
• Are there common approaches for managing large-scale collections?

John Orcutt, Ocean Observatories Initiative [home page]

• linking observational data and users
• uses message-based technology - this enables better enforcement of policy
• uses only cloud services
• has three components (1) deep sea nodes (2) coastal nodes (3) regional cable network
• 25 year operational lifetime
• transformative
• policy-based data systems
• rich data and metadata practices
• scalable architecture
• both geospatial and temporal enabled

Ed Kearns, National Climatic Data Center [PowerPoint presented by Kenneth Galluppi]

Laurin Herr, Cinegrid [PowerPoint]

• media convergence motivates CineGrid
• media technology is driven by 3 sectors - entertainment/media/the arts - science/medicine/education/research - military/intelligence/security/police
• 35mm is still the standard but digital films may never see this format
• hollywood studios now facing decisions about what to throw away - eg out-takes
• use RDF framework for objects
• use iRODS because of programmable microservices

Philip Maechling, Southern California Earthquake Center [PowerPoint]

• data is both observational and from simulations
• the majority of data is from simulations
• 2-way inter-dependence of data and models
• computer-generated models should always have associated verified and validated results and code must be included
• metadata strategies - tools for converting wikis, labels and entries into the metadata store
• advocates practical metadata standard - should be just enough to go from the data to the results

Panel 2 - Promoting Re-use of Scientific Data Collections

Panel Questions:

• How do you handle organization of collections today?
• What are the biggest issues with building collections for new communities?
• When new communities access existing data collections, what new access capabilities are required?
• What level of description is required to meet expectations of new communities?
• Is long-term sustainability enabled through repurposing of collections?
• Are their other driving purposes behind promoting re-use of collections?
• Which institutions can be approached for sustaining re-purposed collections?

Peter Wittenburg, Max Planck Society, Institute for Pyscholinguistics [PowerPoint]

Sudha Ram, iPlant Collaborative [Paper]

• building cyberinfrastructure to support the solution of grand challenges in plant biology
• they are a virtual organisation
• their long term focus is on sustainable food supply
• their challenge is to build an iPlant tree-of-life (IPTOL) - a single tree of evolutionary relationships for all green plants
• methodology: Genotype to Phenotype i.e. construct a methodology whereby an investigator, given the genomic and environmental information about a given plant, can predict it’s characteristics
• modeling is still in it’s infancy compared to other disciplines
• Strong focus on data integration, not simulation: Plant science is truly data driven.
• no single ontology - sequences, gene expressions, whole plant data, environmental and climate data - therefore Data Integration is Key
• issues regarding data integration, provenance, and quality are key enabling technologies
• observes the language barrier between the bio community and the computing/CI community is far higher than any other discipline

Roy Williams, International Virtual Observatory Alliance [PowerPoint]

John Harrison, SHAMAN [PowerPoint Talk, Responses to questions]

Jonathan Crabtree, Odum Social Science Institute [PowerPoint]

Panel 1 Data Life Cycle Management

Panel Questions:

• Which phases of the data life cycle are managed by your repository?
• How do data management requirements differ across the data life cycle?
• What systems do you use to support the data life cycle?
• Can you generalize the mechanisms used to migrate data between different phases of the data life cycle?

Mark McFarland, Texas Digital Repository {PowerPoint}

• TDL has 19 member institutions and use DSpace/Manakin.
• currently moving contents from inside the TDL to a distributed datastore using iRODS.
• still to build a federated discovery layer
• are only just starting to deal with datasets
• collaboration with Mexican research infrastructure

Erin O'Meara, electronic records archivist, Carolina Digital Repository [PowerPoint]

• mission to ensure UNC digital scholarly output is safe and accessible far into the future
• use Fedora as middleware and iRODS for data grid
• content model uses PREMIS/METS/MODS
• custodial shift after objects are deposited
• they apply an OAIS model throughout the architecture
• noted that the data management effort increases during SIP preparation
• their challenge is to develop an approach to active at-risk data - potential approach is precustodial intervention
• CDR Demo

Stacy Kowalczyk, Indiana University, Scholarly Data Services [PowerPoint]

• have a new "Data Publishing service" combining their IUScholarWorks institutional repository and existing Scholarly Data System (was called Massive Data Storage System - MDSS)
• provide data management support and data access
• data will have a persistent URL so it can be linked to publications
• Preservation – bit level
• Use Dspace and for small datasets DSpace bitstreams - for large datasets DSpace records link to stable URLs in MDSS
• facilitate the process for submitters via the DSpace Configurable Submission system
• facilitate the data collection manager’s process via steps in the DSpace workflow system
• engage researchers during projects not at the end
• use immediate benefits to drive automatic capture and 'volunteering’ of metadata
• reduce costs by re-engineering curation processes to leverage this rich metadata and volunteered effort
• Diagram of Data Curation Lifecycle Elements

Cynthia Ghering, Michigan State University [Paper]

• very decentralised environment
• there is NO institutional repository
• Primary Investigator is responsible for curation of their data (chair and dean have oversight)
• they have a digital curation planning project to support investigators

Michael Wright, National Science Digital Library [Paper]

• michael wright was unable to present
• NSDL since 2008 is now a network of repositories for science education materials and not research materials and they have actually de-accessioned much of the research material

Summit Introduction

Desired outcomes were:

• Identification of exemplar data management systems
• Identification of opportunities for federation across institutional repositories
• Identification of interoperability mechanisms at technology, social, and legal levels
• Promotion of collaborations between the identified parties
• Roadmap for the research agenda needed to federate systems.