Overview

This documentation is meant to assist data partners that have contributed data into Oceans 3.0. 

Data Agreements

Formal data agreements are generated for each data partnership in order to establish details like dataset attributions, usage restrictions and more. These are signed by each party, and used as the reference for data stewards when curating metadata records. 

Contact for Data Agreements dataagreements@oceannetworks.ca 

Data Policy and Licences 

ONC's Data Policy is published on our website, and can be accessed here: https://www.oceannetworks.ca/data/data-policy/ 

Part of the ONC Data Policy includes implementing data licences for all datasets in Oceans 3.0, to make it clear what the reuse expectations are. 

Two data licences that we are currently implementing include:

• Creative Commons Attribution 4.0 International (CC-BY-4.0), Constraint: Credit must be given to the creator
• Creative Commons Attribution Non Commercial 4.0 International (CC-BY-NC-4.0), Constraint: Credit must be given to the creator. Only noncommercial uses of the work are permitted

While the above 2 licences are being implemented, there are additional options for third party partners to consider, and select, if they should so choose. These can be found at https://spdx.org/licenses/ 

Licences will be discussed in the scope of Data Agreements going forward. More information about Data Agreements can be found in DataAgreements

Attributions & Credit

Data Citations

See https://wiki.oceannetworks.ca/display/DP/Data+Citations

ISO 19115 Metadata Records

Individual Metadata Record

Metadata, data about data, is information associated with a particular output/product, and is leveraged by data users in search and filtering of repository results. There are several metadata standards in use, and so communities or disciplines define their own best practices; however, there are common metadata elements across all standards, such as author/creator, title, abstract, publication date, etc., that describe the record at a high level. Community or disciplinary metadata standards include more specific information pertaining to that domain, such as ISO 19115 for geospatial science. 

Metadata can be represented in two forms, human- and machine-readable. Human-readable metadata is usually provided to users through a landing page associated with the record; machine-readable metadata underlies human-readable metadata, and is provided in well-established data exchange formats such as XML or JSON. Machine-readable metadata supports automation of tasks; some operations performed using metadata are too large or too tedious to be feasible for humans, and so computers can use machine-readable metadata to talk directly to one another, such as using APIs to harvest records or citations. Not all data users want to sift through records manually, and so APIs require machine-readable records to grab the appropriate results desired by the user.  

Consistent metadata standard usage facilitates interoperability between repositories. Interoperability furthers the implementation of linked data, allows other repositories to harvest records, and increases findability where multiple records exist for the same output/product.

ONC has implemented the latest ISO 19115 standard in our repository, including representing data partner attributions, and linking Research Organization Registry (ROR) identifiers, to ensure that they are properly attributed for tracking their own metrics accurately. The RORs are populated in the Landing Pages through clickable links in the Creators and Contributors sections. Attributions allow ONC to properly identify and credit organizational or personal ownership. We have done this in several places throughout the metadata:

• Creators
• Contributors
• Citation

These attributions are associated with data agreements and roles agreed to by the signing organization. ONC has implemented these attributions across all of our deployments and devices, ensuring contributions are captured appropriately throughout the record and citation text. 

Metadata Catalog

These metadata records are made available through ONC's CKAN Metadata catalogue. From there, they can be harvested for discoverability within other portals or metadata catalogues, such as within the Canadian Integrated Ocean Observatory System Metadata Catalogue. As additional third party harvesters are identified, they will be noted here. 

Roles

The roles associated with each party are restricted to a controlled vocabulary. Our data agreements generally use the same terminology as the ISO 19115 metadata standard, and we have mapped them to equivalent terms in DataCite as follows. 

Research Organization Registry

The Research Organization Registry allocates unique, persistent identifiers to research organizations, much like DOIs for content, or ORCIDs for individual researchers. For example, the ROR ID for ONC is https://ror.org/05gknh003. Over 97,000 ROR IDs have been assigned since the registry launched in January 2019.

Persistent identifiers like ROR IDs ensure that research contributions are correctly attributed, by disambiguating entities which may be known by different names (for instance, 'Ocean Networks Canada', 'Ocean Networks Canada Society', and 'ONC'), or share a name with other entities (the acronym 'ONC' is also used by the 'Office of the National Coordinator for Health Information Technology' and 'Oncolytics Biotech Inc'). When used within the scholarly communications and publication ecosystem, ROR IDs improve discovery, tracking, and linking of research outputs across platforms, organizations, and funders. RORs support the trend towards ensuring credit is given to all parties involved in producing research outputs but have been left out of traditional citations, such as funding bodies.

Your organization may already have a ROR! You can search the registry using your organization's name, URL, or other identifier such as GRID or ISNI.

If your organization is missing from the registry, you can easily request an ID from ROR using their online request form, or email them at info@ror.org.

ROR IDs may be embedded in the metadata for a publication or dataset, or included directly in the citation. When ONC mints a DataCite DOI for a dataset, the attribution metadata is associated with the ROR. In addition, attributions in the ISO 19115 metadata record (within the MD_DataIdentification class) will include the ROR details. If your organization has a ROR ID, but it doesn't appear in the dataset records in Oceans 2.0, please let us know. We'd be happy to update our metadata, and the corresponding DataCite entries. 

Metrics

Digital Object Identifiers (DOIs) are a type of Persistent IDentifier (PID), and make collecting metrics on data use much easier for data collectors and providers. DOIs have been used in the research field for several years, but it has been limited to traditional publications, such as journal articles, until recently. In the past few years the research data community has pushed to have datasets published with DOIs, making tracking usage much more straightforward.

Ocean Networks Canada (ONC) has integrated DOIs into Oceans 2.0 with the MINTED project, funded by the CANARIE Network. DOIs consist of a persistent URL and have several benefits:

• helping users identify the authoritative source easily,
• provides the user with a citation that can be easily 'copy and pasted' into their output, based on the dataset(s) used, making it clear what organizations and/or individual was responsible, and
• allow our data partners to receive credit when their data is reused. 

The research data community has fully adopted the best practice of minting DOIs for datasets, as can be seen in DataCite's statistics: 2019 4.9m total, 2020 8.2m already in April. 

ONC has been following initiatives like the Make Data Count Project closely, to identify the best way to track metrics relating to research data. This project is investigating ways to track views, downloads, and citations, identifying whether current practices would work for data. Through this project, ONC has learned that without a DOI, or PID, associated with a dataset, it is near impossible, and very expensive, to accurately track the metrics associated with a single dataset. These services make use of linked data relationships between PIDs for data, journal publications, researchers and organizations. Thus, including RORs in our DataCite record is an enabler to automatically collecting organization-specific metrics that indicate when datasets are cited within journal publications. 

ONC will be continuing to monitor these developments, with the intention of implementing automated metrics and reporting features in the future. 

Indigenous Data Governance

Ocean Networks Canada is committed to working meaningfully and respectfully with Indigenous and coastal communities. We strive towards meeting the CARE Principles for Indigenous Data Governance. For more information, refer to the wiki page on Indigenous Data Governance.

Pacific Marine Protected Areas Working Group

A metadata services task team is forming to tackle the types of questions stakeholders for marine protected areas (MPAs) are asking:

• understand the impact of human activities in these areas
• transition from Areas of Interest to Marine Protected Areas 
• identify and plan future MPAs and networks of MPAs
• support researchers who have infrastructure nearby
• inform approvals for activities in these regions
• identify crossing of submarine cables

An ONC Science Hub Space has been established to facilitate communications. Feel free to join by using this link. We also have a shared Google Drive to stow relevant materials.

For questions about this topic, please reach out to dfo-mpa@oceannetworks.ca and we will respond promptly.

Contact Information

For clarifications or questions, please contact dataagreements@oceannetworks.ca. Alternately, visit the ONC Help Centre.