Adoption of RDA Outcomes in US Ocean Science Data Repositories

Toward Adoption of RDA Outcomes by US Ocean Science Data Repositories Cynthia Chandler, Bob Arko, Adam Shepherd

US Ocean Science Domain Repositories 2 BCO-DMO Biological and Chemical Oceanography Data Management Office (WHOI) Curation of marine ecosystem system data contributed by NSF funded investigators R2R Rolling Deck to Repository Curation of routine, underway data from US academic fleet, and authoritative expedition catalog Members of Marine Data Harmonization IG

Adoption Process Stages 3 Awareness I just became aware that the output exists Interest I heard about the output, have learned about it, and now have an active interest Evaluation I ve got a strong interest and am willing to commit to evaluating the output Trial I ve evaluated the output, decided I like it, and am willing to give it a try. Adoption It really works!! I ve decided to adopt it and make it part of the system, Yay!

Outcomes of Seven Groups Have Potential for Adoption by Ocean Science Data Repositories 4 Data Citation Data Type Registries PID Information Types Data Foundation & Terminology Practical Policy Metadata Data Publishing (3 of 4 groups)

What does RDA offer domain repositories? 5 BCO-DMO: transition human -> machine clients Ocean science is interdisciplinary Data curation is distributed RDA outcomes can help address challenges of Interoperability between distributed systems in ocean sciences Interoperability between different domains (natural, social) Need solutions that scale for 'Big Data' (VARIETY, VERACITY, velocity, volume) RDA outcomes developed and vetted by representatives from multiple domains

Data Citation (DC) of evolving data 6 DC goals: to create identification mechanisms that: allow us to identify and cite arbitrary views of data, from a single record to an entire data set in a precise, machine-actionable manner allow us to cite and retrieve that data as it existed at a certain point in time, whether the database is static or highly dynamic DC outcomes: 13 recommendations and associated documentation ensuring that data are stored in a versioned and timestamped manner identifying data sets by storing and assigning persistent identifiers (PIDs) to timestamped queries that can be re-executed against the timestamped data store

Description of Data Citation Outputs 7 Data Versioning: For retrieving earlier states of datasets the data needs to be versioned. Markers shall indicate inserts, updates and deletes of data in the database. Data Timestamping: Ensure that operations on data are timestamped, i.e. any additions, deletions are marked with a timestamp. Data Identification: The data used shall be identified via a PID pointing to a time-stamped query, resolving to a landing page.

Adoption of Data Citation Outputs 8 Evaluation Evaluate recommendations Try implementation in existing systems Trial BCO-DMO: R1-11 fit well with current architecture; R12 doable; test as part of DataONE node membership R2R: curation of original field data and selected subset of post-field products (ship track); so no evolving data Both working with DataONE as our aggregation system and service provider

Data Type Registry (DTR) 9 DTR mission: see if it is possible to make implicit assumptions about data contained in datasets explicit and programmatically share these assumptions using types and type registries DTR outcomes: 1 website and 1 API. The registry website provides a user interface for someone to describe both simple and complex data types used for data within a project. They can also search data types created by others. The API provides a way to programmatically interact with the registry including the ability to import data type descriptions.

Description of DTR Outputs 10 1. Data Type Registry A website with a GUI that provides a way for an authorized someone to describe a data types used in data products. 2. Data Type Registry API An API that among other things creates JSON representations of the information about data types. This is a pointer to the API specification implemented in the data type registry mentioned above.

Adoption of DTR Outputs 11 Evaluation evaluate the registry and API for use in existing data repositories try using the prototype registry to record a set of data types and then provide some example code that uses the API to access the information in the data type registry Trial BCO-DMO, R2R, GeoLink? data type determined by instrument type R2R already maintains a de facto library of file types for environmental sensor systems in the US research fleet, in collaboration with NCEI and Chronopolis. We could publish this as a formal Data Type Registry

PID Information Type (PID-IT) 12 PID-IT goal: Provide a way to harmonize PID information types (and associated information) that are associated PID across disciplines and PID providers. (Also to provide technical solutions) PID-IT persistent outcomes: Types for example use- cases have been registered in the type registry developed in this WG

Description of PID-IT Outputs 13 1. Type Examples and Illustration Use Cases These are examples of API Description A description of the API used to access the PID registry created by this group API Prototype Implementation A working version of the API connected to the PID registry that has been created Registry Prototype The registry prototype itself Client demonstrator GUI Demonstration of the registry and it s use via a graphical user interface developed by the group s intern. 2. 3. 4. 5.

Adoption of PID-IT Outputs 14 Evaluation evaluate the client registry GUI Trial BCO-DMO and R2R PID systems in use: DOI (datasets and expeditions), ORCID, FundRef (US awards), ISNI (global organization), IGSN (global samples), re3data (repository), and domain-specific for instruments and measurements Possible: R2R, BCO-DMO and NCEI (US ocean archive) joining DataONE; DataONE architecture is well-aligned with PID-IT approach, so perhaps DataONE adopts the PID-IT API and offers that as a service to the community

Data Foundations and Terminology (DFT) 15 DFT mission: to understand what the core of the RDA data domain is and then develop definitions of core terms based on data models. This effort is a part of the effort to form agreement on RDA culture. DFT persistent outcomes: 4 Documents and 1 Wiki Tool that summarize the work they have done on Terminology. The wiki tool is intended to be used by other RDA WGs and IGs to extend the terminology terms, etc. beyond those determined to date.

Description of DFT Outputs 16 Overview Document model descriptions Report on the discussion about a large number of data models Analysis & Synthesis Document Report on the analysis of the data models considered by the group Term Snapshot Document Report on a snapshot of core terms that have been identified Use Cases (1), (2) Use cases that describe how other working groups use the terms the group has been talked about Semantic Media Wiki Term Tool Tool to capture initial list of terms and definitions for DFT WG discussions, open for others to use. (it is kind of persistent at this point) Report of Interactions w others about terms Summary of conversations with ~120 individuals about data in the context of DFT findings. 1. 2. 3. 4. 5. 6.

Adoption of DFT Outputs 17 Evaluation evaluate core terms in Semantic Media Wiki Term Tool for ocean science domain as they extend our current term reference source Trial BCO-DMO: maplocal system terms to the DFT terms; adding deployment type (cruise, dive, float, experiment), R2R: already essentially follows this model; publishes Collections that represent a field expedition (research cruise), having Persistent Identifier (DOI). Challenges Need dereferenceable term URIs (DTR ?) Does the Semantic Wiki Tool provide relationships or a way to describe them? Is each term a concept in an ontology (e.g. OWL file)? Governance?

Opportunity . . . 18 Interaction of DFT, PID-IT, DDRI and DTR ?? Registering terms/types at DTR (federated system of distributed, production ready, registries) What is the appropriate level for a registry? Professional, domain-specific societies? (ASLO, AGU, EGU) Institutional library? Community organization (ESIP, OGC, ODIP) ? Not having operational registries may hinder adoption

Practical Policy (PP) 19 PP goals: To enable sharing, revising, adapting, and re- using of computer actionable policies for sharing data, particularly in a data repository and to suggest a set of generic policies to be applied to our data; collect and register practical policies PP persistent outcomes: Practical Policy (PP) WG recommendation package of Policy Examples and Template Workbooks

Description of Practical Policy Outputs 20 1. Policy Template Template that includes a generic set of policies suggesting how they can be implemented within a data system. Implementations Policy descriptions and implementation details 2.

Adoption of Practical Policy Outputs 21 Evaluation evaluate the policy template and implementation documents Trial Identify policies that should be documented BCO-DMO and R2R are DOI Publication Agents, but not long-term archives. Consider documenting our practices (for archive and replication) in a computer-actionable format, so data deposits (to NCEI, Chronopolis, DataONE) can be periodically verified as part of a self-audit process

Metadata 22 Metadata group goals: Set up a sustainable, community-driven RDA Metadata Standards Directory, designed for users rather than automated tools, that provides brief details for common research data. Compile a set of use cases that analyze and document the various ways in which metadata can be used (e.g. for discovery, exchange, re- use, etc.) Metadata group outcomes: UK DCC Disciplinary Metadata Standards Catalogue http://www.dcc.ac.uk/resources/metadata-standards functional GitHub prototype directory with version control http://rd-alliance.github.io/metadata-directory/

Adoption of Metadata Group Outputs 23 Evaluation Current Status BCO-DMO: ISO 19115-2 (19139 compliant), DIF, CF via NVS, O&M, PROV, RDF, DCAT, Dublin Core, (OAI-ORE soon), (used to do FGDC, but dropped that recently in favor of ISO-19115) Compare DCC lists with current practices Identify standards where we currently have none Identify mismatches and consider addressing them R2R: all of the above plus DataCite and IGSN for samples

Publishing Data Working Groups 24 Publishing Data Workflows Publishing Data Services Publishing Data Bibliometrics Evaluate and Monitor: R2R and BCO-DMO will evaluate these outcomes and look for ways to implement in our current architecture, or relevant communities (promoting recommended practices) GeoLink (NSF EarthCube) using Linked Data (Semantic Web technologies) to connect data and publications meaningful data use statistics would be very welcome

Infrastructure Components 25 supporting the modern research endeavor, is like creating a quilt; a work of art created by a community of practitioners with a shared goal. Each member of the community lovingly, and laboriously designs and constructs their piece of the quilt.

Creating the Infrastructure for a Domain 26 Putting the pieces together to create the whole block for a domain.

RDA 27 Combining the domains, And adding the unifying framework, to create the global research data quilt (RDA)

Thank you!

29 Others (6 other group outcomes available by Sep 2015) Data Description Registry Interoperability (DDRI) Infrastructure providers & data librarians to find connections across research data registries and create global views of research data. Repository Audit and Certification DSA WDS A convergent DSA-WDS certification standard will help eliminate duplication of effort, increase certification procedure, coherence and compatibility thus benefitting researchers, data managers, librarians and scientific communities. RDA/WDS Publishing Data Bibliometrics RDA/WDS Publishing Data Services universal interlinking service between data and scientific literature RDA/WDS Publishing Data Workflows Wheat Interoperability

SLIDE TEMPLATE: 3 slides per working group 30

Group name (GRP) 31 GRP goal: Description GRP outcomes: Description

Description of GRP Outputs 32 1. output output description 2. description

Adoption of GRP Outputs 33 Suggestions evaluate try using the product Test cases BCO-DMO R2R GeoLink