NIH | National Cancer Institute | NCI Wiki  

The objective of the Semantic Requirements page is to solicit requirements and comments regarding the next generation semantic infrastructure used by caBIG® and NCI CBIIT.

The semantic infrastructure used currently by caBIG® and CBIIT has evolved over a dozen years. The progress that we have made in learning how to use data semantics to advance sharing of research data and analysis tools have led to increased expectations regarding the role of semantics in translational research. These expectations, together with the rapidly accelerating exchange of information between basic research and clinical practice domains, will require a new generation of semantics infrastructure and new approaches for using semantics.

We envision establishing new operational procedures and tools, etc. that will support computable semantic interoperability (CSI). CSI is the exchange of <<non-ambiguous meaning>> between machines. CSI includes both static semantics (the meaning of data describing "things") and dynamic, behavioural semantics (information used to describe service and system behaviors, interactions, etc.). We expect that CSI, the next-generation semantic infrastructure, will need to support a variety of broad semantic requirements including:

  • straight-forward data discovery and queries;
  • analytical scenarios involving both data discovery and subsequent processing by multiple applications;
  • negotiation of service contracts using service meta-data, functional profiles and semantic bindings;
  • support for formal representation of "context" including relationships and cardinalities.

During the month of May 2009 users were invited to tell us about high-level use cases and requirements that they would like to see "next generation" of the caBIG/CBIIT "semantic infrastructure" meet. In early June a number of Project Initiation Proposals were developed to address these requirements. These proposals were planned to constitute the bulk of the semantics work program for the coming year or two. They would be in briefed to the various domain workgroups and submitted to the C-Team for prioritization, approval/rejection, and funding.

The Vocabulary Knowledge Center wiki and forums were available to provide us the with requirements that you have for semantics support. Information was sought on what has worked well in our current infrastructure, and what has not worked well, and what you need that is not supported at all by our current semantics and structure (including EVS, caDSR, GME, Grid advertising and discovery).

We requested your feedback across the board:

  • Lessons you have learned about creating or using data semantics
  • Needs for less time-consuming or less complicated ways of creating data and service semantics
  • Needs for new types of semantic information driven by your scientific or clinical work
  • Ideas for making it easier to create or to use semantic information
  • Comments, suggestions, criticisms, etc.

COMMENT PERIOD IS NOW CLOSED. Please visit the Vocabulary Knowledge Center Semantic Requirements Forum to REVIEW use cases, comments and suggestions.

  • No labels