Skip to end of metadata
Go to start of metadata

Meeting Information

Date:

February 21, 2008

Time:

11 AM EST/4 PM GMT

Attending: JB McGee, co-chair; Susan Albright, Ben Azan, Jeroen Donkers, Dennis Glenn, Mike Hagan, Debbie Sher, Valerie Smothers, Dan Walker

Agenda Items

1 Review exemplar VP package (most recent player and data specs)
Ben explained that he developed a valid scorm virtual patient package. The structure of the manifest is important. There are schema files that the LMS needs to be able to use the package, and some examples of actual XML that validate. The data specification does list all of the files that are needed in a VP SCORM package.

JB asked Ben to review SCORM and why we are using it. Ben explained we wanted VPs to be valid as SCORM packages so that they can be played in a Learning Management System. We need to make sure the package validates. JB asked if SCORM usage was to be written into the spec. Ben commented it will become part of the spec, and developers creating their own VP packages can look at the example package for assistance. JB asked if tools to validate were available. Ben commented that Reload and Tool Tracker were available. Valerie added that an ADL conformance test suite exists, and that Nabil is developing a conformance suite as part of eViP. Ben added there are some errors in reload coming from files from scorm. He's getting more information on that. JB asked how it would work without a player. Ben made a dummy player as place holder.

JB asked Ben what he needs from the working group. Ben commented that testing would be helpful. Any feedback is welcome. Debbie commented the VA has access to an LMS. That would be a good test. Susan asked if Blackboard supports scorm. Valerie commented that she thought so, but she wasn't sure which version. Ben included a javascript file to facilitate tracking in an lms; he is not sure if it will communicate properly. Susan commented they could try it in educommons.

Ben will update the package with modified files. Ben will try the adl conformance test suite, too.

Mike asked about the data specification. In a number of elements there are source and source id attributes to reference a vocabulary. How would the specification handle post - coordinated terms? Valerie asked Mike to explain what post-coordinated terms are. Mike explained that there are 400,000 concepts in snomed. Even with that rich availability, there are things that don't exist, and users have to concatenate concepts to get what they want. For example, you may need an adjective to modify hypertension. Will that cause difficulties with how id is used by the player? JB commented that it depends on the player. The specification doesn't go to implementation level for vocabulary terms. He commented it would be out of scope for the specification; a better player would be able to handle such terms. JB asked if it occurs often. Mike replied it does. Debbie commented they do that with their digital library, too. Mike asked about diagnosis likelihood. Is there a sense of what those mean? Is High greater that 75 percent chance? Susan commented that from their perspective, there was no percentage thought of. It is up to individual users to determine. JB agreed. In their VPs, authors wanted to give feedback on certain actions.

Mike added that in paragraph 6.2.2, there is a typo, launchable "wrapper" is misspelled. Also, under medication, there is an example of myocardial infarction, with appropriateness set to "never," which is clinically inappropriate.

Jeroen asked a question related to the activity model. How do we indicate the first node? Ben commented that first node listed in the file would be used as the first node in the activity. We'll double check to make sure that information is in both the player and data specifications.

2 Revisit metadata location within manifest
Valerie summarized that in the last call we discussed the placement of metadata within the manifest. Metadata can be placed under the root element of the manifest, or it can be placed at the organization level. Schawn Thropp of ADL recommended putting metadata at the organization level, and Rachel had asked for staff to see how other systems handle metadata and gather more expert opinions, including that of the Technical Steering Committee. The TSC concurred with Schawn Thropp's recommendation to put metadata at the organization level. Valerie asked Dan if he had looked into how educommons imports metadata. Dan hasn't looked at educommons to see how it handles metadata. Dan commented that educommons does not read metadata at the manifest level. They look at the resource level. He does not think they look at the organization level. He will talk to them to find out more. JB commented we should go with the organization level unless there are comments.

3 Revisit XHTML implementation
JB asked Valerie to summarize the discussion on the last call and current progress. Valerie explained that Rachel had relayed some problems using XHTML on the last call. The Labyrinth interface allows users to put their own html markup tags in the text, but the tags are not well formed XHTML. As a result, they don't validate. To get around this, she used URL encoding to encode the VPDText. Valerie and Ben discussed the situation with the Technical Steering Committee, and they recommended using an XHTML parser to detect errors and allow users to correct them. Valerie asked Jeroen for his opnion since there had been some problems with XHTML at Maastricht as well.

Jeroen commented that on the last call we had discussed whether we should allow the full xhtml or a subset. That is already in the specification, but specific tags are not listed. It is a good solution to force authors to use xhtml, and systems can transfer html to xhtml. Susan commented that some way to help authors would be good. Are there wysiwyg editors for xhtml? Dan commented the current editor is not xhtml. There may be some faculty that don't use wysiwyg , and it could be tricky getting them on the right path. Dan ran his own encoding solution. He thought that might be a good solution. Ben questioned the usefulness of encoding VPDText, which is effectively sidestepping the standard. JB added the point is to get the data into a format that is easily reused. Susan found there is an xhtml wysiwyg editor called x standard. Dan commented a minority of users would enter their own tags. Anyone could make an error there. JB commented a good authoring system would check for things like that. How much do we want to force good habits and good design? Jeroen agreed. Ben asked f we would still be restricting what xhtml tags are used. JB said that was reasonable.

Dan asked if there was a list of allowable tags. Ben said he could take a stab at it. The group recommended bold, italics, underline, bullet lists, and tables. Dan asked if we were using wiki notation. Valerie commented that it was taken out so that xhtml could be passed directly to the browser for display. Dan asked about the image tag in xhtml. The current spec says not to use it, but many users have a paragraph of text and a media resource, or they put images in a table. Mike commented they do that a lot in their testing materials. Dan commented that in the general course packaging world, there is precedence for direct referencing of images in html pages referenced by course packages. JB agreed we should put it on the agenda for next time.

4 Triggering nodes
Dan commented that he would like to discuss when dam node refers to other dam nodes, can embedded dam nodes use the trigger attribute? Can a dam be triggered from another dam node? They have natural hierarchy in the subtest. The group agreed to look at it on the next call, and Valerie asked Dan if they could speak about this between calls.

5 eViP update
Valerie related that the eVip application profile has been updated to reflect the Feb 11, 0.46 version of the standard. The eViP group has a technical meeting April 9, and they will have more of an update after that meeting.

6 Virtual Patient meetings and sessions at annual conference
Valerie summarized the virtual patient sessions on at the annual conference. Tuesday there is a Virtual Patients in a Nutshell session, followed by a Virtual patient authoring workshop, developers meeting, and open working group meeting. Wednesday there are several virtual patient demonstrations, and Thursday there is a virtual patient panel. JB commented that after the developers meeting, it would be good to get wider input from the virtual patient community from authors and users.

7 Virtual Patient Poster at annual meeting (slice poster, template)
Valerie explained that the reception this year would include an exhibit hall and posters on each of the working groups. This would allow attendees to learn more about the working groups and meet those involved prior to the working group meetings the next day. Valerie posted a slice of life poster that could serve as the basis for this new poster. JB commented it would benefit from updates, including what we find out from the eViP meeting in April. Many will want to know when is something coming out. Susan commented we should have working group members listed as other working groups have. Valerie offered to merge the slice of life poster with the template and come up with a hybrid. JB said the diagrams are still valid. Jeroen commented it would be good to show the same patient in two different players. Tufts has one that they shared with NYU and Pitt. Jeroen added that in eViP they will have translated VPs. JB agreed we should reference that exists. He recommended filling in what next with examples. JB has pictures he will send to Valerie.

Decisions

  • Metadata will be placed at the organization level of the manifest barring any further comments.
  • Recommended xhtml tags will be incorporated into the specification.

Action Items

  • Debbie will test running the VP package in the VA LMS
  • Susan will test running the VP package in Blackboard (if compatible) and educommons.
  • Ben will update the exemplar package with updated files.
  • Ben will try validating the package using the ADL conformance test suite.
  • Ben will correct the typos noted.
  • The group will discuss integrating text and images for the next call.
  • The group will discuss triggering on the next call.
  • Valerie and Ben will set up a call with Dan to discuss triggering.
  • Valerie will draft a virtual patient poster for the next call.
  • No labels