Draft Minutes: ITANA call of 13-Mar-2014
Rich Stevenson, University of Maryland University College (host)
Chris Eagle, University of Michigan
Leo Fernig, University of British Columbia
Scott Fullerton, University of Wisconsin - Madison
Piet Niederhausen, University of Washington
Ashish Pandit, UC San Diego
Jim Phelps, University of Washington
Brenda Reeb, University of Rochester
Jon Terrones, University of Wisconsin - Madison
ITANA Website: http://itana.org/
ITANA Wiki: https://spaces.at.internet2.edu/display/itana/Home
Standard Data Schemas
Ashish led a discussion of standard data schemas in higher education. At UCSD Ashish is working on defining web services schemas as an overall effort to replace legacy integration methods. He investigated the School Interoperability Framework (SIF), which attempts to cover K-12, postsecondary, higher ed, and adult education and provides XML schemas, as well as Common Education Data Standards (CEDS), which provides a logical model.
Ashish asked the group who has used standards-based model for web services design, and what their general strategy is. Our discussion included:
Leo mentioned Post Education Secondary Council (PESC) standard for college and high school transcript exchanges, which is widely used across North America. Another group of standards are the IMS Global standards for product interoperability between, for example, and student system and a learning management system.
Scott described the experience at UW-Madison, which has tried to stay close to the IMS Global standards for student and person data. Most of the work was driven by specific needs on initiatives such as advising.
Jon added that in middleware, UW-Madison has combined the HR-XML and IMS Global specs into a schema to represent person objects for directory services, using as much of the predefined standards as possible and only adding a few domain-specific elements.
Jim described a strategy of building off standards as much as possible, in the hopes that vendors will eventually meet those standards in interfaces.
Brenda described an approach starting from metadata and governance. Last year Rochester launched a business-driven data dictionary, endorsed by data stewards. The business definitions are linked to the physical locations of the data. Rochester uses Data Cookbook with a heavy emphasis on governance, and also considered Collibra. The governance-driven approach encourages IT to develop strong relationships with business units and raises everyone’s ability to talk about data.
Rich observed that K-12 has a higher adoption rate of some of the standards than higher ed, perhaps because they are forced to share data more, with more rigorous reporting requirements from regulators. Also, K-12 does not create as many of its own software solutions. For higher ed, there are only a few major vendors and this may have limited the need for interoperability and the need for standards. There are overlapping standards that continue to evolve but with limited adoption.
Scott described UW-Madison’s ERPs as a barrier to adopting standards, because of a cultural perception that standards could not add much to what PeopleSoft provided — even though integration was increasingly difficult. This culture changed once there was a need to consume data from diverse administrative systems — for example, LMS with needs for curricular data. A historical exception to this was person data for identity management purposes.
Brenda observed an increase in the number of smaller scale (especially cloud-based) solutions with a need for especially person data. There is a lack of standards for creating feeds for these solutions implemented outside central IT.
Leo described European initiatives to standardize transfer credits and asked about activity in North America. Jim indicated that transfer credits usually involve a manual evaluation based on factors such as the type of institution the student is coming from — in the absence of a standardized curriculum. Rich mentioned that there are state-level efforts to create cross-walks. There are some transcript service providers that are looking to adopt standards to get better data from institutions. Scott added that another long-term driver can be increased emphasis on outcomes-based education, which requires standardization.
Architecture Value Scorecard
Jim gave an overview of an architecture value scorecard he is developing at the University of Washington. The scorecard tries to value the impact of projects based on architectural value in addition to business value, so that initiative owners and governance groups can increase alignment with architectural goals.
The scorecard covers business alignment, user impact/usability, information/data architecture, technical architecture, and risk management. Based on 9 principles the scorecard assesses 17 values. It bridges between general principles and specific future-state roadmaps.
The documents Jim shared are available here:
In a next iteration, the scorecard will also link to services that a project should be taking advantage of to be in alignment. In an initial review of about 60 projects using the draft scorecard, there were projects that not every part of the scorecard would apply to.
The scorecard can be used throughout the gating process as a project is proposed and becomes real. Although the scorecard won’t be a basis for stopping a project, it can help assess the cost/debt that organization is taking on if the project scores lower in some areas. That kind of debt is more appropriate for some types of projects (innovative or differentiating projects) than others.
In terms of governance, Jim described his plans to apply the scorecard to the project review board and to expanding the enterprise architecture steering group to include participants who can speak to each area on the scorecard.
Next call: March 27, 2014.