In last few years the International Organization for Standardization (ISO) 19115 family of metadata standards has become the predominantly accepted worldwide standard for sharing information about the availability and usability of scientific datasets among researchers. US interests in the ISO standard have also been growing as global-scale science demands participation with the broader international community. Yet, due to the complexity and rigor of the ISO metadata standards, adoption has been slow. In addition, support for data content in current implementations of the standard has been minimal.
In 2009 the Alaska Data Integration working group members (ADIwg) mobilized to jointly address the common data integration efforts. Beginning in 2012, ADIwg started to focus on difficulties associated with generating and exchanging ISO 19115-2 metadata among its partners which vary in size, from small NGO research groups to state offices, universities, and large federal bureaus. This differential in size and technical capabilities among its membership brings with it a diversity in metadata requirements and the ability to provide the necessary technical support to its researchers. The ISO 19115 metadata standard is not a single standard but a family of ISO and GML (Geography Markup Language Encoding Standard published and maintained by the Open Geospatial Consortium, Inc.) standards. Understanding this amalgamation of standards to a degree sufficient to generate a valid metadata record requires an investment of time well beyond what should be reasonably expected of any PI or end user. These issues are not specific to ADIwg and its members; they are shared by many organizations striving to transition to the ISO metadata standard.
After much discussion, the ADIwg membership agreed to co-author a software toolkit with an architecture that has the flexibility to achieve the following goals:
- Isolate the complexity, rigor, formatting, and terminology of the ISO standards from the user. Users should not be required to know anything about the ISO standards.
- Provide clear developer and user documentation.
- Design a layered architecture so developers will have access to each layer. Developers should be able to enter the toolkit at whatever layer best fits their needs; e.g. download the code library to do some custom programming or post JSON metadata to a hosted JSON to ISO translator.
- Design the architecture to accommodate incremental development cycles. Add features without disrupting previous implementations.
- Implement all code and services as open-source software projects using a GitHub repository.
- Provide flexibility to write to multiple metadata standards, including ISO 19115-2:2009, 19115- 1:2014, 19110, and possibly FGDC.
- Host a publicly available version of the online metadata editor for end users to enter and edit metadata then request a valid ISO standard metadata record to be returned.
- All code should be written in a popular computer language, be platform independent, royalty-free, and available through an open-source repository to encourage participation from and benefit to the widest possible user base.
To be clear, these tools are for the generation metadata only. This project is not intended to provide services as a metadata clearinghouse, metadata storage, or search and modify capabilities.