NMFS Enterprise Data Management 1 DAARWG Meeting December 8, 2010, 2010 Jim Sargent, NOAA Fisheries...
-
Upload
louisa-carol-mathews -
Category
Documents
-
view
215 -
download
2
Transcript of NMFS Enterprise Data Management 1 DAARWG Meeting December 8, 2010, 2010 Jim Sargent, NOAA Fisheries...
NMFS Enterprise Data Management
NMFS Enterprise Data Management
1
DAARWG MeetingDecember 8, 2010, 2010
Jim Sargent, NOAA FisheriesInformation Architect
NMFS Enterprise Data Management
The End in Mind: Take Aways
• Need to make quality data available to the world • It will take a cultural shift - Which is HARD!• NMFS EDM: One LO’s approach
• Understand the comprehensive enterprise-wide process we have been through, and
• Possibly leverage the fruits of our labors and learn from challenges • DAARWG recommendations:
– Expand scope beyond Access and Archive Requirements to full Data Management Life Cycle
– Promote and support EDMC and LO data management efforts– Leverage similarities while respecting diversity of data– Review and support EDMC’s Procedural Directives– Consider:
• Adopting a new vision of data management• Doing a comprehensive inventory
NMFS Enterprise Data Management
Outline
• A Case for a New Vision for Data and Data Management
• One LO’s approach: NMFS EDM
• Recommendations
NMFS Enterprise Data Management
• Volume and diversity of data• Collected for one-use; potential for multi-use• Infrastructure needed to support multiple access mechanisms• Releasability of data (organizational vs. technical) • Coordination of collection and distribution• Comparison and utilization of modeling outputs and observations
Overall Need for:• Clear and consistent data management policy• Better data documentation (metadata)• An overarching response plan
• Need to quickly identify applicable / available data
Deep Water Horizon Data Management Challenges
Prepared by Environmental Data Management Committee for NOAA Leadership
5
Pre and Post Disaster
Immediate Response
NMFS Enterprise Data Management
“NOAA is like a library
without any card catalog
…or even bookshelves”
……..Dr. John A. Knauss, Former NOAA Administrator, 1986
NMFS Enterprise Data Management
• President's Directive on Open Government – Transparency, participation, and collaboration– Timely publication of quality information
• It’s the right thing to do!
• Greater than archive and preservation issues– Not all access is done from archives
NMFS Enterprise Data Management
Everyone is Wresting With This
• Other federal agencies – IWGDD, NSF, ICES, ISO
• DM maturing as a discipline– DAMA DM Book of Knowledge (DMBOK)– Certified Data Management Professionals
• H. R. 5037 April 15, 2010– Requires Federal agencies to develop public access
policies
• NMFS Science Center Data Management Reviews
NMFS Enterprise Data Management
It’s About a Cultural Shift• Need to move from the academic paradigm
– Publish or Parish Share or Parish– Science Science Information– Science Quality good enuf
• Changing Culture is Hard– Tipping Point, Blink
…. Malcolm Gladwell
– Switch, Made to Stick … Phil and Dan Heath
NMFS Enterprise Data Management
A New Vision??
NOAA data assets are recognized and managed as a core agency resource, on par with financial and human resources.
11
NMFS Enterprise Data Management
Science Board
Leadership
13
• Analysis & Research•Recommendations:
• Policies• Procedural directives• Guidelines• Best practices
• Implementation• Marketing
Approvals and support• Policies; • Procedural directives• Guidelines• Best practicesNMFS Staff
FIMAC
NMFS Staff
NMFS Enterprise Data Management14
2008 2009 2010
Jan 15, 2009IA Position
Filled
FIMC researched NMFS DM and developed
EDM Recommendations
• Recommendations presented to and accepted by LC
• FIMAC Created
•Teams Created• IMCs Created
FIMAC Workshop• Teams’ Plans developed• Required Resources Identified
Policy drafted and
vetted
Data Documentation
Procedure Directivedrafted and vetted
DataInventory Initiated
ImplementationPlanning and
Development of TeamsResearch, Analysis, and
Recommendations
A Brief History of EDM Time
Data Stewardship Teams Established
NMFS Data Catalog
established
Policy Enacted
NMFS Enterprise Data Management
15
NMFS EDM MISSION:
“To effect a cultural change in which all NMFS data are recognized and managed as a core agency resource, on par with financial and human resources.”
NMFS Enterprise Data Management
16
NMFS EDM Vision
NMFS customers
can confidently find, access,
and use our data
NMFS Enterprise Data Management
17
NMFS EDM Vision
NMFS customers
can confidently find, access,
and use our data
Internal and external constituents
NMFS Enterprise Data Management
18
NMFS EDM Vision
NMFS customers
can confidently find, access,
and use our data
Confidence in finding and trust in using our data
NMFS Enterprise Data Management
19
NMFS EDM Vision
NMFS customers
can confidently find, access,
and use our data
• Using various portals, data.gov, geospatial.gov, etc) • Browse through an ordered hierarchies or taxonomonies• Search using:
• Discipline specific key words (controlled vocabularies)• User tags (folksonomies)• Metadata include users’ comments
• Minimize the number mouse clicks
NMFS Enterprise Data Management
20
NMFS EDM Vision
NMFS customers
can confidently find, access,
and use our data
Through confidentiality and security filters while using standard tools and formats
NMFS Enterprise Data Management
21
NMFS EDM Vision
NMFS customers
can confidently find, access,
and use our data
Download selected data withsufficient documentation, including quality indicators and warnings to effectively and properly use and understand the data
NMFS Enterprise Data Management
22
NMFS EDM Vision
NMFS customers
can confidently find, access,
and use our data
All NMFS enterprise data we choose to share
NMFS Enterprise Data Management
23
NMFS EDM Vision
NMFS customers
can confidently find, access,
and use our data
NMFS Enterprise Data Management
24
NMFS Enterprise Data And Information Management Policy
Overview
• Enacted by Eric Schwaab: June 2010– Directed RAs, SDs, and ODs to send 2-3 people to NMFS
Data Stewardship Workshop• General Policy: All data shall:
– be visible, accessible and understandable to authorized users;– modeled, named and defined consistently across and within
all NMFS programs;– have a standard set of metadata; and– be managed, controlled and shared by data stewards
throughout the data management lifecycle. – be publicly available generally within one-year of its collection
A high level NMFS Policy Directive implemented by Operational Procedure Directives
NMFS Enterprise Data Management
25
Procedural DirectivesPrinciples/Concepts
• Developing Procedural Directives an iterative process
• Data Documentation Procedural Directive: FMCs shall:– Develop their own plans for documenting and sharing data
assets– Inventory and document data assets and tools in NMFs
metadata repository, InPort– Measure metadata quality with a rubric
FY11 will be a year of learning and practicing how to document and share our data
• Develop, use, and refine metadata standards• Populate InPort with discovery level metadata• Develop practice and refine procedures for sharing
NMFS Enterprise Data Management
NMFS Data Stewardship Workshop
• Discuss New Data Management Policies and procedures
• Develop best practices– How to document our data
• e.g., quality assessment
– Spirals in data documentation
– Rubrics and metrics
• Q&A and Sidebar Communities
27
To develop a shared understanding of data stewardship, empower the data stewardship teams to lead the way for implementing the Data
Documentation Procedure Directive, to develop best practice based on the experience during the initial documentation and inventory tasks.
NMFS Enterprise Data Management
Next Steps
• Data Doc. PD approved – Dec 2010
• Data Inventory Complete – Dec 2010
• Data Doc. Implementation Plans Due- Feb 2011
• Data Stewardship Workshop – Apr 2011
• First cut Best Practices – Apr 2011
• Data Doc. Implementation Plans Finalized – May 2011
• Address Preservation and Terminology
Socialize, Support, Promote, and Market
NMFS Enterprise Data Management
Critical Success Factors
• Management Commitment• Dedicated, Personally Committed Team
– FIMAC– Coordination Team– EDM Partners– Steady Hand at the Helm– Executive Sponsorship
• Drive Up and Then Drive Down– Harness the mavens
• Socialize, Support, Promote, and Market 29
NMFS Enterprise Data Management
Challenges
• Tipping Point almost reached but not there yet
• Budget Shortfall– Perception that field not fully bought into it
• Overworked Teams a high risk
NMFS Enterprise Data Management
DAARWG Recommendations• Expand scope beyond Access and Archive
Requirements to full Data Management life Cycle• Support cultural change needed to meet demand
for quality data• Promote and Support EDMC and LO DM Efforts• Review and Support Procedural Directives• Leverage similarities while understanding
diversity– Establish good data management practices for the whole
data management lifecycle• Consider Recommending
– A new Vision of Data Management – Conducting a comprehensive inventory with defined
metadata
NMFS Enterprise Data Management36
Data Sharing
• Data shall be shared in data.gov, as appropriate, as a one-click data asset, a one-click product, or by using a software tool (e.g., FOSS)
• Sufficient documentation to understand the data being shared must be published in InPort and referred to or provided with the data
• Data Stewards decide what data is appropriate for sharing
• Sharing confidential data must conform to Agency policy
“Data should be made as widely and freely available as possible while safeguarding the privacy of participants,
and protecting confidential and proprietary data”
NMFS Enterprise Data Management
37
Timelines for Sharing Types of DataData Asset
Type Description Sharing Time Frame
Continuous Time Series
Continuous data collections for the purpose of monitoring fisheries or environment
Annually, according to FMC Procedural directives implementation plan
Discrete Time Series
Surveys that collect data at discrete time intervals across months or years
Periodically based on survey design, according to FMC implementation plan
One-time collections
Results of experiments or small one-time collections
1 year or upon publishing of a work based on the data asset
Multiple Source
Compilation of data from multiple sources, e.g., compendiums syntheses, scientific reviews and studies
Upon publishing of a work based on the data asset, according to FMC implementation plan / publication policy
Derived Data*
Data and information developed using statistical and mathematical models and other technics.
1 year or upon publishing of a work or decision based on the data asset
Does not include provisional, predecisional documents and preliminary analyses leading to final management actions
NMFS Enterprise Data Management
1. Data should be archived and accessible
2. Adequate resources for end-to-end management
3. Management activities should involve users
4. Interagency and international partnerships
5. Metadata are essential
6. Expert stewards required for management
7. Process to decide what data to archive
8. Archive must support discovery, access, and integration
9. Effective management requires a formal, ongoing planning process
Prepared by Environmental Data Management Committee for NOAA Leadership38
National Research Council Committee on Archiving and Accessing Environmental and Geospatial Data at NOAA, 2007
Principles for Effective Environmental Data Management
NMFS Enterprise Data Management
39
Identifying the Problem
• Top priority Issues – Critical gaps – No authoritative data inventory– Insufficient metadata– Data quality and consistency challenges– Ability to integrate data– Administrative systems
• Other issues– Data being lost– Data not being archived for perpetuity– Historical data that need rescuing – Communications re: applications and IT
The FIMC identified 12 key issues and interviewed their management to determine their priorities
IssuesIdentification
Interestingly, the
lowest ranking
issue was that NMFS did not have buy-in
across FMCs for
addressing IMreturn
NMFS Enterprise Data Management
40
Identifying the Problem
• Top priority Issues – Critical gaps – No authoritative data inventory– Insufficient metadata– Data quality and consistency challenges– Ability to integrate data– Administrative systems
• Other issues– Data being lost– Data not being archived for perpetuity– Historical data that need rescuing – Communications re: applications and IT
The FIMC identified 12 key issues and interviewed their management to determine their priorities
IssuesIdentification
Interestingly, the
lowest ranking
issue was that NMFS did not have buy-in
across FMCs for
addressing IMreturn