This section includes the following:
Overall Process and Request for Community Input
This is the process as published on the NCI Wiki.
INCLUDE the published process here.
Goals and Objectives for the Semantic Infrastructure 2.0 Roadmap
The purpose of this document is to define the basic requirements for the next generation – that is, Version 2.0 – of caBIG®-CBIIT Semantic Infrastructure.
"Semantic Infrastructure" is defined as including the collective set of strategies, technologies, and tools that support the publishing and processing of all aspects of data, metadata, terminology, and value set content related to the support for both design-time and run-time computable semantic interoperability (CSI).
Thus it is assumed that Semantic Infrastructure 2.0 subsumes the scope of the current caBIG®-CBIIT semantic infrastructure including caDSR and EVS. In particular, the Semantic Infrastructure 2.0 is expected to support relevant aspects of both informational ("static") and behaviorial ("dynamic") metadata in addition to historical terminology and value-set management functions.
It is especially important to note that the support for informational, behaviorial, and terminology and value-set management will be integrated with the caGrid 2.0 infrastructure to allow for both design-time and run-time support for CSI at the level of individual services deployed on caGrid 2.0.
Critical aspects of the overarching framework for integration of the Semantic Infrastructure 2.0 with the caGRID 2.0 technical infrastructure are expected to emerge during evolution of both this document and the caGrid 2.0 Roadmap. Ultimately these aspects will be harmonized and unified in a single technical approach. The integration will be significantly deeper than the current integration as supported by the Global Message Index (GMI).
The Semantic Infrastructure 2.0 Roadmap is being developed in a transparent, collaborative process with the caBIG® community to ensure that the relevant requirements needed to provide the planned design- and run-time support for informational and behavioral semantics integrated with the caGrid 2.0 infrastructure are appropriately surfaced and defined. The overarching enterprise-level requirement for both the Semantic Infrastructure and caGrid 2.0 projects is to provide comprehensive support for the information discovery and integration, and functional coordination requirements, that will meet the evolving needs of the scientists, clinicians, patients and other stakeholders who collectively define the caBIG® community.
It should be emphasized that although the Semantic Infrastructure 2.0 and caGrid 2.0 roadmaps are being developed by separate project teams, it is the overarching goal and commitment of CBIIT that the two efforts ultimately be deeply integrated and merged into a single caBIG® 2.0 semantically-aware Service-Oriented Architecture and run-time platform, that is, caGrid 2.0.
Thus, both the Semantic Infrastructure 2.0 and caGrid 2.0 roadmap projects share the same overarching goals:
- Reduce the barrier to entry for all users of the Semantic Infrastructure 2.0 and caGrid 2.0
- Provide a "linear value proposition," that is, make "easy things easy to do while hard things will require more effort." This means providing tools and solutions that can be used and assembled by multiple stakeholders from multiple perspectives as opposed to the current high-level of effort required to "get on the Grid," a effort which has, in fact, proven to be a barrier to participation for many stakeholders in the caBIG® community.
- Provide support for legacy users and their data.
- Provide an environment that utilizes informatics standards when appropriate to enable broad-based intra- and inter-community interoperability, but does not necessarily enforce standards when they are not needed due to considerations of limited scope or minimal interoperability requirements.
The Semantic Infrastructure 2.0 Roadmap document will form the basis for subsequent requests for proposals that will result in development work that will realize the goals defined by the roadmap. In particular, the Semantic Infrastructure 2.0 project is expected to produce both the roadmap document in close collaboration with the caBIG® community, and to develop a comprehensive set of prototype tools. These tools will inform follow-on development work in terms of:
- Definition and design and run-time support for informational and behavioral metadata
- Appropriate adoption of emerging Semantic Web tools and technologies
- Support for migration of legacy semantic infrastructure data and metadata
- Integration strategies and tooling requirements for interoperability with non-CBIIT semantic representational strategies and infrastructures
Thus readers of the Semantic Infrastructure 2.0 Roadmap should view it as a "Scope and Vision" document executed as part of a formal "Semantic Infrastructure 2.0 Inception Phase." Thus, future versions of the roadmap document can be expected to provide a concrete overview of topics including, but not necessarily limited to, the following:
- Major risk factors (including but not limited to complexity, tooling, and scalability) that must be managed or mitigated to enable the Semantic Infrastructure 2.0 project to succeed
- Description of prototyping activities
- Identification of various "build versus buy" options for specific components of the Semantic Infrastructure 2.0
The open and transparent framework in which the Semantic Infrastructure 2.0 project is being executed is specifically designed to encourage and integrate caBIG® community input into the project's overall content. This includes integrating caBIG® stakeholder community input into the specifics of the several bulleted items in the preceding paragraphs.
In the style of the caGrid Roadmap, the following are the "big buckets" of capabilities which collectively define the Semantic Infrastructure 2.0:
In summary, the overarching philosophy behind the content of the Semantic Infrastructure 2.0 Roadmap is that the technologies and approaches described are a direct result of the need to support the requirements of the scientists, clinicians, trialists, patients, and other stakeholders that comprise the caBIG® stakeholder community. Thus the following two sections of this roadmap provide a working profile of the stakeholders and the working set of associated use cases and storyboards that are collectively being used to define the overarching, enterprise-level requirements for the Semantic Infrastructure 2.0 technologies and tools.
The Semantic Infrastructure 2.0 Roadmap is being collectively authored by a team with extensive knowledge of the legacy caBIG® semantic infrastructure and the associated lessons learned from its use, as well as the realities of "what works and what doesn't" in the technology world as it moves to "Web 2.0" and "the semantic web." As described in the published process framework for the development of the Semantic Infrastructure 2.0 Roadmap, the goal of the project is to combine the perspectives and inputs of the Semantic Infrastructure 2.0 team of experts with caBIG® community input to provide the content for a roadmap that will guide the development of the next generation of caBIG® semantic infrastructure.
Reference Frameworks for Development of Semantic Infrastructure 2.0
With respect to the overarching technology strategies being adopted to guide the development Semantic Infrastructure 2.0, there are several "givens:"
Support for legacy semantics as currently represented in caDSR via ISO 11179: It is critical that the semantics currently represented in caDSR and EVS be supported in the Semantic Infrastructure 2.0. However, it is equally important that both current and future caBIG® stakeholders understand that as users of the Semantic Infrastructure 2.0, the underlying representation of legacy semantics may – and in fact will – be markedly different in the Semantic Infrastructure 2.0 compared to the legacy semantic infrastructure. One of the primary responsibilities of the Semantic Infrastructure 2.0 Roadmap is to specifically identify and describe prototypes for strategies and tools which will be required to both migrate existing semantic content into the Semantic Infrastructure 2.0 as well as support external semantic infrastructures which may be based on representational approaches that differ from those being adopted by the Semantic Infrastructure 2.0.
Semantically-Aware Service-Oriented Architecture (sSOA)
Both the Semantic Infrastructure 2.0 and caGrid 2.0 will be developed and deployed within the context of an overarching approach to enterprise architecture which uses the distributed computing design paradigm commonly referred to as Service-Oriented Architecture (SOA). In addition, because of the fundamental importance of semantics in any architecture approach in the context of the life sciences and healthcare (in the broadest sense of those terms), the SOA being developed by NCI CBIIT as manifested in the Semantic Infrastructure 2.0 and caGRID 2.0, is referred to as a "semantically-aware SOA" (sSOA). It is beyond the scope of this document to discuss in detail the various benefits and goals, core organizing motivations, or fundamental design principles of SOA. However, the following bullet points summarize each of these topics. Interested readers can refer to a number of references including two texts by Thomas Erl: "Principles of Service Design" and "SOA Design Patterns."
Organizing Principles of SOA
• Business-driven
• Vendor-neutral
• Enterprise-centric
• Composition-centric
Benefits and Goals of SOA
• Intrinsic interoperability
• Increased federation
• Increased business and technology alignment
• Increased vendor-diversification options
• Increased IT ROI
• Decreased IT burden
• Increased organizational agility
SOA Design Principles
• Standard Service Contracts
• Service Loose Coupling
• Service Abstraction
• Service Reusability
• Service Autonomy
• Service Statelessness
• Service Discoverability
• Service Composability
Services-Aware Interoperability Framework (SAIF)
The Semantic Infrastructure 2.0 is in large part the operational support for the metadata defined the CBIIT implementation guide for the HL7 Services-Aware Interoperability Framework (SAIF). Readers interested in the specifics of the metadata defined by the CBIIT SAIF implementation guide should consult that document directly http://nci.xxx.xxx. In particular, the chapter on the Enterprise Conformance and Compliance Framework (ECCF) provides focal point for the definition and representation of the collective set of informational and behavioral metadata which the Semantic Infrastructure 2.0 will support at both design- and run-time (via the caGrid 2.0 platform.)
Relationship of the Semantic Infrastructure 2.0 to caGrid 2.0.
The project to define the Roadmap for version 2 of the caBIG® semantic infrastructure, that is, the "next generation" of the current caDSR and EVS semantic infrastructure is being conducted in parallel with and is fundamentally coordinated with the caGrid 2.0 Roadmap project. The ultimate goal of the two efforts is to produce two complementary but integrated Roadmaps that will enable caGrid 2.0 to provide expanded semantic processing capabilities (when required) including but not limited to run-time service discovery and resolution of semantic queries. Details of both the Semantic Infrastructure 2.0 Roadmap and the caGrid 2.0 Roadmap, be publicly available and open for community input.
Expectations and Guidelines for Review and Feedback
This is the first working draft and is expected to evolve with each release.
