Summary: IODE/JCOMM Forum on Oceanographic Data Management and Exchange Standards

Description of Activity

The IODE/JCOMM Forum on Oceanographic Data Management and Exchange Standards, hereafter called the IODE/JCOMM Metadata Meeting, was held 21-25 January, 2008 at the IOC Project Office for IODE in Oostende, Belgium. The stated objective of this and future meetings is "to get broad agreement and commitment to adopt a few standards related to ocean data management and exchange, thereby assembling not only the agreed conditions by which we operate, but in fact enact these agreements in our respective organizations." This group is establishing an international framework and endorsement body for metadata standards, information exchange practices, and quality control practices, under the auspices of IODE/JCOMM.

Key topics for the meeting included:

• Metadata,
• Vocabularies, Code Lists, and Ontologies,
• Quality Control, and
• Standards Endorsement Process.

Participation was by invitation only, in an attempt to maximize the efficiency and progress of the first meeting. (18 people attended, including Peter Pissierssons and Robert Gelfeld who weren't listed.) Some discussion took place about how to engage the larger community, though this topic was by no means exhausted.

Results

Official results of the meeting will be distributed, so this summary is brief and high level.

Overall, the meeting was extremely productive. A lot of material was introduced and discussed, and areas of agreement and further work were identified. In many cases, a wide variety of resources were consolidated as never before, setting the stage for further progress.

The meeting targeted high-level strategic objectives and activities, focusing an opportunities and needs in the JCOMM, IODE, and international data center communities. While much of the work, and many of the conclusions, will be applicable to all marine science data activities, the meeting particularly considered overarching structural strategies and solutions. These will be documented in activities and reports resulting from the meeting.

Metadata Standards

The metadata standard ISO-19115 was discussed, as was the marine community profile building upon it. Other metadata standards were also mentioned. [I don't recall that any definitive outcomes directly addressed these standards. - ed.]

This topic led to discussions about processes for selecting and endorsing metadata standards (see below).

Vocabularies, Code Lists, and Ontologies (and Formats)

Ontologies

After a moderately lengthy presentation on ontology resources, the group discussed likely directions for this technology. The participants familiar with the current state of ontology developments considered the technology on the brink of widespread adoption, likely to be heavily used in 2 to 5 years.

A significant topic that arose in both the ontologies presentation, and several vocabulary discussions, was the need for infrastructure to support the use of interoperable ontologies and vocabularies.

Also, ontologies were often cited as key resources toward achieving vocabulary integration.

Vocabularies and Code Lists

The forum considered:

• The entities to be described by the vocabularies (parameters, instruments, ships etc.) with particular reference to the requirements for the marine metadata profile,
• The vocabularies covering these entities that are already in existence,
• The infrastructure for vocabulary content governance (making the decisions concerning vocabulary content), and
• Vocabulary technical governance (vocabulary maintenance and access through systems such as the NERC DataGrid vocabulary server.

Vocabulary domains included units, platform types, platform (read: ship) instances, science keywords, parameters, sensor types, and biological taxa. In some domains, it appeared one or two vocabularies were dominant, though recommendations for needed work were usually identified.

Many of the vocabulary discussions spoke to the infrastructure required to register, manage, and use vocabularies. Jianping Mao of Global Change Master Directory, and Roy Lowry of BODC/SeaDataNet, both presented material representing significant enhancements to existing vocabulary work available at their sites.

Discussions also related to what makes a good vocabulary. John Graybeal presented a template ('quick and dirty') matrix developed for his Instruments Vocabularies presentation, which several other presenters reused. This topic is likely to receive considerable attention at MMI and elsewhere in coming months.

Formatting Considerations

Formatting considerations for the specific metadata types date and time, latitude, longitude, and altitude was also discussed. Best practices were suggested in each of these areas, and will be published. (The scope over which the best practices should apply -- metadata or data, in what content standards and projects, etc.) was not fully resolved, but will be addressed as materials are prepared.)

Metadata Endorsement Practices

The participants identified a need for an endorsement process for standards for IODE/JCOMM projects. This endorsement process takes into account processes in use at NASA and IOOS, and will be further developed and presented for consideration (and eventually approval) by the IODE/JCOMM community.

A lot of developments directly impact MMI activities and planning.

Standards Evaluation Criteria: The MMI standards evaluation criteria and vocabularies evaluation matrix will inevitably be developed further, inside and outside of MMI. (IOOS is likely to address the same topics shortly, if not already.)

Vocabulary Integration: Comprehensive vocabulary mapping is also possible in some areas, thanks to the research performed for this activity. This could be a lead-in activity to developing comprehensive ontologies in similar areas. Candidate domains include platform and sensor types, and possibly some subset of science keywords.

Ontology Development: In addition to the vocabulary integration developments, work being performed at NASA's Global Change Master Directory, SeaDataNet, and MMI are reaching levels of maturity that suggest collaborative work on ontologies may be fruitful.

Best Practices Reference: The endorsement process being discussed in this meeting is fairly rigorous and lengthy. Meanwhile, technical developers around IOOS and the world want specific recommendations in all the areas described here, and more. There is room for a process that highlights key standards as being worthy of particular consideration when creating your own data system; MMI is getting very close to being able to create this process.

References: MMI has already added over a hundred references in preparation for this workshop, and reorganized its Vocabularies, Tools, and Organizations sections to make finding appropriate references easier. Additional references from this work, and from the OceanTeacher web site, will be added. In some cases (e.g., ontology tools) MMI must seriously evaluate whether to duplicate references elsewhere on the web.

Ocean Teacher: MMI must investigate all possible collaborations with IODE's Ocean Teacher project, to maximize beneficial outcomes while minimizing costs. The educational/training tools may provide an appropriate home for MMI training materials.

Endorsement Processes: These are sprouting vigorously; MMI needs to track them and make sure all are aware of the others.

Funding Opportunities: MMI was specifically asked for a white paper on one of these topics, and specific proposals could be considered to perform several others.