Child pages
  • MACE-Dir BOF notes of April 2014
Skip to end of metadata
Go to start of metadata

Notes: MACE-Dir BOF of Tues. April 8, 2014 at Global Summit in Denver


Keith Hazelton, UW-Madison (Co-Chair)
Tom Barton, University of Chicago
Mark Scheible, MCNC
Mike Danay, Fremont NE , representing the ESU Coordinating Council
Jordan Clark, Educational Service Unit, Nebraska
Scott Isaacson, NE, ESU, coordinator
Jim Peterson, University of Illinois at Urbana-Champaign
Jason Radford, University of Illinois at Urbana-Champaign
Bernic A'cs, University of Illinois at Urbana-Champaign
Paul Riddle, UMBC
Roland Hedberg, Umea University, Sweden
Albert Wu, UCLA
Heather Flanagan, Spherical Cow Consulting
Benn Oshrin, Spherical Cow Consulting
Emily Eisbruch, Internet2, scribe

MACE-Directories Working Group
MACE-Directories Wiki:


Schema Needs for K12

Keith: Representatives from the Quilt InCommon Pilots ( ) have reached out to the MACE-Directories Working Group to discuss schema needs for K12. it's important to start with concrete projects and initiatives and describe and document those. Then we can determine what attributes are needed to support those actual use cases.

A template for sharing use cases is here:

Use case from MCNC

Important Questions:
-For federation at the university level, the university IDP often has the needed info. In a K12 scenario, the source of info may be a local district and there may be a need for a call to the service record, which can be more complicated.
-What are the needs of actual service providers and who is the authoritative source for the info that must get asserted?
-Where will the IDP run?
-Where do users authenticate?
-Who issues the credentials?

Nebraska Quilt InCommon Pilot

The Nebraska pilot team is hoping to develop a structure to serve at some level for all K12 in the state, serving educators, students, and parents.
The pilot in development involves an Ed-Fi data dashboard for 9 school districts in Nebraska.
This may work to collect various sources of data from Info systems and present that to administrators and teachers to look at students in aggregate and individually.
Talking to the Ed-Fi people has been a large project. The data architects are working through all of that. We start with the identity side. Have a unique identifier for staff and students. Hope to be able to answer questions such as: what courses is this teacher responsible for & what assessment data do we have for this student?

Q: that sounds like the whole educational data store being available to this provider. Are they creating local user accounts at the resource?
A: It's a hybrid model
-some accounts are stored onsite at school districts
-some accounts are delegated to ESU's (Educational Service Units)

Q: There is one large LDAP environment?
Jordan: Yes we have 2 consortium areas and rest of the state is loosely affiliated.

Q: what's driving what you are doing versus the state Dept of Ed is doing?
A: We provide the identity side of the dashboard, The state decides what to do with the data.
We look at schools as ultimate source of identity info. Dept of Ed is the aggregator of data. We in the middle are providing services

Q: What about course enrollment data?
A: We centrally hold that within our directory info.

Need to provide authentication mechanism for parents,
There are some issues there.

Illinois Quilt InCommon Pilot

Illinois has much in common with the Nebraska pilot story. All K12 will have similar issues.

For details on the Illinois pilot, see

Setting values re Entitlements

Best to have SPs communicate what data they will need.

Q: Are there items that Illinois and Nebraska can work on jointly around this, or are the projects in such different stages?

A: we don't want every pilot, every effot to be a one-off

Within Ed-Fi, there is a common thread in how the data is defined.
Other important standards?

US Dept of Education has outlined Data Standards and this fits in.
Important to get in the loop of the Dept of Ed. Work in this area.

MarkS: we approached this same issue of a schema last year. We still need to develop the use cases that have specific requirements.
Then we can see if its covered by the eduPerson model. Maybe we are not at that point yet.

There may be a need to indicate if a person falls under COPPA restrictions.
We want to provide grade level.
We need the use cases

Bernie: we can articulate a couple of common use cases, such as the use of Student Rosters to are support applications

JimP: we have 30 differnet apps to which we need to provide roster info. Different formats for each one. Must wait until day 10 of school. Then the classroom teachers must do it themselvers for updates.

The roster handling for teachers is particularly burdensome In a high school, since the teacher has multiple groups of students.

Keith: there is some mix, there is system of record , data driven rosters for being in a certain class.

Bernie: In a K12 environment we can assume they are using the student info system. Data maintained by office staff.
It may be delegated or centralized. The challenge is to propagate the student roster info in a standard way that 's consumable by most vendors.

It is automatable with some basic agreement between vendors.

Q: Does eduPerson support such a notion of a class roster?
A: no
Mark: it gets back to how much the application will do. In the North Carolina IDM service, the student records have an association with the teachers. The connection happens in the student record. A service translates the number into the teacher name. That is within the IAM infrastructure.

Bernie: That's why we look at two layers. We have the 4 eduPerson schema attributes to identify a user. Beyond that, we must include subdomain to represent buildings.

Next Steps: Gather concrete use cases and document them at


ORCD registry of Researcher Identifiers

Keith: the MACE-Directories Working Group is monitoring work around ORCD Identifiers.

One model is being used at Penn State: any incoming faculty member, grad student or researcher will be assigned an ORCD identifier. This is done in the backend by the IDM and this will be persisted in the institutional repository. That ORCD identiter is used to correlate all output pertaining to that researcher.

Another use case is around SciENcv:

SciENcv would like to be able to use the ORCD identifier for researchers and this identifier will be used when applying for a grant.

Email List

To join the MACE-Directories Email List:

  • No labels