This wiki has undergone a migration to Confluence found Here
<meta name="googlebot" content="noindex">

Difference between revisions of "HL7 Tooling Challenge"

From HL7Wiki
Jump to navigation Jump to search
Line 1: Line 1:
This page is to record the ideas to create an HL7 Tooling Challenge similar to the Government Challenge initiative - see examples under Health at: [http://www.Challenge.gov Government Challenge]
+
This page is to record information about the first HL7 Tooling Challenge.  The intent of this initial challenge is to produce tooling to help implementers implement V3 standards which are expressed in Model Interchange Format.   
   
 
The theme requested was to produce tooling to help implementers implement V3 standards which are expressed in Model Interchange Format.   
 
  
 
*  The Model Interchange Format is organized as a set of XML Schema files that describe the key metadata to express published standards.  The MIF is used to transform the approved models as serialized XML files.  The MIF was designed to meet the requirements of the V3 Methodology to express concepts that could not be expressed in XMI.
 
*  The Model Interchange Format is organized as a set of XML Schema files that describe the key metadata to express published standards.  The MIF is used to transform the approved models as serialized XML files.  The MIF was designed to meet the requirements of the V3 Methodology to express concepts that could not be expressed in XMI.
Line 7: Line 5:
 
* ''' The following ideas were proposed - we haven't decided which of these ideas will be proposed, although we have concluded that the second proposal should be postponed as it is dependent on the first idea.'''
 
* ''' The following ideas were proposed - we haven't decided which of these ideas will be proposed, although we have concluded that the second proposal should be postponed as it is dependent on the first idea.'''
 
*# Produce a UML Profile for MIF Static Models - This is a task so that other modeling tools have a chance to be configured to understand MIF
 
*# Produce a UML Profile for MIF Static Models - This is a task so that other modeling tools have a chance to be configured to understand MIF
*# Produce a MIF to XMI XML transform consistent with an updated UML Profile - This task is dependent on the completion of task 1.
+
**Document a UML Profile for the semantics in the MIF representation of the RIM and other derived models - identifying those concepts that cannot be expressed in UML Profile language to provide to OMG for possible extension of UML
 +
**Using the resulting UML Profile, adapt Enterprise Architect to express the proper HL7 stereotypes as a proof of concept. An example would be enumerations for the drop down stereotypes for a concept domain, code system or value set. A further example would be a class could be an Act, Entity, or Role
 +
 
 +
 
 +
*  The criteria on which the submissions will be judged will be:
 +
** The proposed UML profile will be reviewed by an OMG representative to ensure it's validity
 +
** The proposed UML profile is successfully applied to Enterprise Architect such that EA can produce class models that can represent the semantics described in the UML profile - documenting any difficulty in applying the profile.
 +
*  The AWARD will be given to the first team submitting a validated UML profile and corresponding UML Profile applied to Enterprise Architect. Announcement of the successful team is desired to be at the September 2012 Working Group during the plenary. If no team submits a successful solution the contest will continue.
 +
 
 +
*The following tooling objectives would also add value, and will be used to solicit existing or contributed tools - and potentially be the subject of a different HL7 Tooling Challenge in the future.
 +
*# Produce a MIF to XMI XML transform consistent with the updated UML Profile.
 
*# Produce a viewer that can browse MIF schemas (Meta level)- this task makes it easier to understand the relationship among the MIF files and their elements.
 
*# Produce a viewer that can browse MIF schemas (Meta level)- this task makes it easier to understand the relationship among the MIF files and their elements.
 
*# Produce a viewer that can browse MIF based models (model level)- this task makes it easier to understand the serialized models expressed in MIF XML.
 
*# Produce a viewer that can browse MIF based models (model level)- this task makes it easier to understand the serialized models expressed in MIF XML.
 
*# Produce a viewer that can browse instances from MIF based models (instance level) - this task makes it easier to evaluate instances based on serialized MIF models.(there is some work that already exists)
 
*# Produce a viewer that can browse instances from MIF based models (instance level) - this task makes it easier to evaluate instances based on serialized MIF models.(there is some work that already exists)
 
* A poll remained open until January 5.  An invite was sent out to the Co-chairs list to participate and help identify which tasks fulfill burning needs. No responses were received.  This initiative will be marketed during the San Antonio WG and another poll will be send out after that solicit input.
 
 
*  The Tooling Work Group will describe the criteria on which the submissions will be judged once a challenge has been determined.
 
*  The submission due date will also be identified would the submissions be due - we need to be realistic about how long a challenge might take.
 
 
Please put your suggestions on this page - the Tooling WG will review suggestions on the Tooling call Feb. 2, 2012.
 
 
Additional suggestions from WGM in San Antonio:
 
*Assign the proper the proper HL7 stereotypes to the UML artifacts in Enterprise Architect. An example would be enumerations for the drop down stereotypes for a concept domain, code system or value set. A further example would be a class could be an Act, Entity, or Role
 
 
*From RIMBAA, which has its own tools page at [[Tools for RIM based software development]]: #2 is certainly a key issue for software developers. We already have an example of #3 as an educational tool made available by a commercial party, #5 is probably already (mostly) avialble as the Instance Editor. The MIF based 'evaluation' of instances, without validating the instance, doesn't sound terribly attractive.
 
*From RIMBAA, which has its own tools page at [[Tools for RIM based software development]]: #2 is certainly a key issue for software developers. We already have an example of #3 as an educational tool made available by a commercial party, #5 is probably already (mostly) avialble as the Instance Editor. The MIF based 'evaluation' of instances, without validating the instance, doesn't sound terribly attractive.

Revision as of 16:01, 9 February 2012

This page is to record information about the first HL7 Tooling Challenge. The intent of this initial challenge is to produce tooling to help implementers implement V3 standards which are expressed in Model Interchange Format.

  • The Model Interchange Format is organized as a set of XML Schema files that describe the key metadata to express published standards. The MIF is used to transform the approved models as serialized XML files. The MIF was designed to meet the requirements of the V3 Methodology to express concepts that could not be expressed in XMI.
    • Requirements for MIF[1]
  • The following ideas were proposed - we haven't decided which of these ideas will be proposed, although we have concluded that the second proposal should be postponed as it is dependent on the first idea.
    1. Produce a UML Profile for MIF Static Models - This is a task so that other modeling tools have a chance to be configured to understand MIF
    • Document a UML Profile for the semantics in the MIF representation of the RIM and other derived models - identifying those concepts that cannot be expressed in UML Profile language to provide to OMG for possible extension of UML
    • Using the resulting UML Profile, adapt Enterprise Architect to express the proper HL7 stereotypes as a proof of concept. An example would be enumerations for the drop down stereotypes for a concept domain, code system or value set. A further example would be a class could be an Act, Entity, or Role


  • The criteria on which the submissions will be judged will be:
    • The proposed UML profile will be reviewed by an OMG representative to ensure it's validity
    • The proposed UML profile is successfully applied to Enterprise Architect such that EA can produce class models that can represent the semantics described in the UML profile - documenting any difficulty in applying the profile.
  • The AWARD will be given to the first team submitting a validated UML profile and corresponding UML Profile applied to Enterprise Architect. Announcement of the successful team is desired to be at the September 2012 Working Group during the plenary. If no team submits a successful solution the contest will continue.
  • The following tooling objectives would also add value, and will be used to solicit existing or contributed tools - and potentially be the subject of a different HL7 Tooling Challenge in the future.
    1. Produce a MIF to XMI XML transform consistent with the updated UML Profile.
    2. Produce a viewer that can browse MIF schemas (Meta level)- this task makes it easier to understand the relationship among the MIF files and their elements.
    3. Produce a viewer that can browse MIF based models (model level)- this task makes it easier to understand the serialized models expressed in MIF XML.
    4. Produce a viewer that can browse instances from MIF based models (instance level) - this task makes it easier to evaluate instances based on serialized MIF models.(there is some work that already exists)
  • From RIMBAA, which has its own tools page at Tools for RIM based software development: #2 is certainly a key issue for software developers. We already have an example of #3 as an educational tool made available by a commercial party, #5 is probably already (mostly) avialble as the Instance Editor. The MIF based 'evaluation' of instances, without validating the instance, doesn't sound terribly attractive.