2nd Stakeholder workshop: Bertin, Embrapa's appraoch to open Agricultural Science
1. Embrapa’s approach to Open Agricultural Science
Patricia R. Bello Bertin
Secretariat of Management and Institutional Development
Brazilian Agricultural Research Corporation – Embrapa
e-ROSA 2nd Stakeholder Workshop | Wageningen, 27-28 November 2017
2. I. Introducing ‘Embrapa’
- Founded in April, 1973
- Public, Research & Development organization (Brazilian Ministry of
Agriculture, Livestock, and Food Supply)
www.embrapa.br
3. - Founded on April 26, 1973
- Public research organization (Brazilian Ministry of
Agriculture, Livestock, and Food Supply)
www.embrapa.br
I. Introducing ‘Embrapa’
6. • Goal: integrating existing initiatives into a common framework
(e-infrastructure)
II. Embrapa and e-ROSA’s aproach: convergence
Source: Report of the First e-ROSA Stakeholder Workshop (6-7 July 2017, Montpellier)
7. III. Current efforts and developments
• Raising awareness; community engagement
• Mapping out key stakeholders and existing initiatives (inventory
and assessment of existing services)
• Articulating and connecting data sources and infrastructures into a
common framework at the institutional level (distributed e-
infrastructure; focus on interoperability, data reusability and
exchange) → federated dataset catalogue (trusted data
repositories)
• Collaboratively developing a common vision and an institutional
strategy towards Open Science
8. IV. Data landscape
• Distributed, heterogeneous resources: a variety of data sources;
custom-built, in-house systems which are not integrated and not
easily accessible; content relevance not yet been evaluated;
semantic resources poorly explored (no discovery services).
• Limited data sharing (mainly amongst co-workers and partners).
• Individual storage resources (e.g. workstation and external
devices).
• Tension between business interests and societal priorities
(transparency and access to data).
1. UnFAIR
2. Findable, Usable for Humans
3. FAIR metadata
4. FAIR data, restricted access
5. FAIR data, Open Access
6. FAIR data, Open Access, Functionally Linked
Source: Report of the First e-ROSA Stakeholder Workshop
(6-7 July 2017, Montpellier)
9. Motivations:
1. Transparency in public management
2. Society’s contribution with innovative services to the citizens
3. Quality improvement of government data
4. Enabling innovative businesses
5. In compliance with the law!
V. Open Science and Open Government
• Open government data: “data that is produced, collected or kept under
the custody of public authorities, which is disseminated under Open
Data schemes and terms of use”
Source: Union Accounts Tribunal, Brazil (available at: http://portal.tcu.gov.br/biblioteca-digital/cinco-motivos-para-a-
abertura-de-dados-na-administracao-publica.htm )
10. Marco
regulatório
Dados
Abertos
VI. Open Data legal and regulatory framework
DADOS ABERTOS
Public consultation by the Federal Government:
databases of interest to the scientific community?
National Infrastructure of Spatial Data
Decree n. 6.666/2008
Law of Access to Information
Law n. 12.517/2011
National Infrastructure for Open Data
Normative Instruction n. 4/2012
Open Data Policy of the Federal Executive Branch
Decree n. 8.777/2016
Acessing and interlinking governmental databases
Decree n. 8.789/2016
11. • If a database is not classified as ‘sensitive’ (legal hypotheses of industrial
secrecy, justice secrecy, intellectual property, etc.), it should be made open
→ created a process for requesting access to governmental databases.
Highlights from Decree 8.777/2016
• Requires the creation of ‘Open Data Plans’ by governmental entities
(strategic document that guides the implementation and promotion of
open data schemes, in accordance to minimum quality standards, so as to
facilitate its understanding and reuse).
12. Institutional project: ‘Inserting Embrapa's technological solutions
(assets) into the digital business market‘ (2017)
VII. Building up Embrapa’s Open Data Plan
OPEN DATA
Two representatives per
Research Unit identified
(Multidisciplinary team - III)
ODP execution
(Implementation - VII)
ODP revision
(Updating - VIII)
Open Data Plan (ODP)
coordinator nominated
(Leadership - II)
ODP creation and online
publication
(Dissemination - VI)
Governmental Open Data Policy
presented to the Board of Directors
(Sponsorship - I)
Workshops and webconferences
(Conceptual alignment - IV)
Data sources inventory
(Collaborative process - V)
13. Guiding question: From all data produced by your Research Unit,
which datasets possess greater potential for openness?
• Relevance to the society/segments of public (Citizen Service System)
• Data that is already public
• Legal obligation or commitment for data dissemination
• Data referring to the strategic actions of the Research Unit
(specificity)
• Maturity level in terms of data organization (existing databases)
Prioritization criteria: (High: 3; Medium: 2; Low: 1)
VII. Building up Embrapa’s Open Data Plan
14. • Difficulties with distinguishing public data from sensitive data
(competitive advantage or risk to R&D projects). → institutional policy?
VIII. Difficulties and lessons learned
• Types of data most commonly associated to the notion of open data:
» 1st Administrative data (management
acts, contracts, personnel, etc.)
» 4th Climatologic data
» 5th Biodiversity data
• Secondary, aggregated, and processed data are more easily associated
with the notion of open data than primary data.
» 2nd Socio-economic data
» 3rd Geospatial data
» 6th Soil data
• Maturity in data organization: data with potential for openness are not
always structured, and may be in possession of the researcher and not the
organization (or not even exist in a digital format). → Data rescue,
documentation, management and curation are a prerequisite for opennes
and reuse.
15. • An audience-specific communication plan is needed in order to overcome
cultural barriers: managers, researchers, ICT professionals, librarians, etc.
• Brazilian research funding agencies called to support Research Data
Management and Open Science.
VIII. Difficulties and lessons learned
• It is important to value and strengthen existing initiatives and to involve all
organizational areas in the construction of the ODP (Research, Strategic
Management, Business, Technology Transfer, Communication, IT, etc.).
• Collaborating with government initiatives might provide mutual support
and greater visibility of actions.
• It is necessary to achieve an understanding of RDM practices of different
scientific cultures represented at the organization, in order to define
parameters and descriptors that guarantee greater quality in data
organization and recovery.
16. • Useful to speak of the ‘open spectrum’, clarifying the internal public
about the apparent dichotomy between the opening of data and the
processes aimed at guaranteeing information security, confidentiality and
intellectual protection.
• Data opening should be carefully thought and planned as part of an
institutional program, demand-oriented and aligned to the corporate
governance.
VIII. Difficulties and lessons learned
• New competencies, skills and habilities are required (data librarian, data
engineer, data scientist, etc.).
• Analytic capacilities need to be further developed (the ‘data provider’
perspective, on its own, is not culturally accepted).
17.
18. IX. How the international community could help
• Disseminate high impact use cases, as well as the variety of existing
resources/infrastructures.
• Incentive mechanisms are required at the institutional level in order to
encourage researchers to make their data FAIR/Open).
• Develop methodologies for data ‘FAIRification’ and indicators (metrics) to
facilitate monitoring and encourage data opening.
• Need for shared legal framework and guidance: intellectual property
issues, data ownership and related licences.
• Develop sustainable businesses models.
19. Obrigada!
“The important thing in science is not so much to obtain new facts
as to discover new ways of thinking about them.”
William Bragg, quoted in Reif and Larkin (1991, p.739)
patricia.bertin@embrapa.br
Supervisor for Transparency and Data Governance
Secretarial of Management and Institutional Development
+55 (61) 3448-1808
Editor's Notes
Population: 200 million people
Wide, geographically-distributed research organization