1 of 30

Consortial Hyku for Open Resources

Gretchen Gueguen

PALCI

gretchen@palci.org

2 of 30

Consortia

Hyku

OER

3 of 30

OER

Consortia

Hyku

4 of 30

Open Educational Resources (OER)

5 of 30

Fun Fact: College is Expensive

Source: Florida Virtual Campus, 2018 Student Textbook and Course Materials Survey

6 of 30

Textbooks Affect Student Learning

Source: Florida Virtual Campus, 2018 Student Textbook and Course Materials Survey

7 of 30

OER Community

OER = Open (openly licensed) Educational Resource

Textbooks
Quizzes and other assessments
Classroom activities
Complete online courses

Open license = free to use, adapt, and remix �(in some cases)

8 of 30

Why not OER?

Source: Opening the Textbook: Educational Resources in U.S. Higher Education, 2017

9 of 30

Gaps in OER Repository Support

Publishing and peer review workflows
Remix and adaptation support
Post-publication review and rating
Support for interactive objects like quizzes and courses

Vendor hosted software isn’t flexible enough

Support isn’t built into Open Source options

But there are bigs gaps in support for these types of materials in traditional repositories.

Firstly, we aren’t traditionally publishers and don’t necessarily have skills or tools in house to support the peer review process and the editorial workflows. Some libraries have gotten involved in open journal publishing, and there are similar needs there, but there are differences too.
A big advantage of OER is that faculty can take what’s in a textbook and tailor to their course, but they need some support for figuring out how to do the remixing and adaptation. In repositories this could be tools to be able to actually edit the materials, explanatory help with licensing, opportunities for re-publishing, etc.
A big deal for faculty is the usage of a book and the review of it by other faculty, so a lot of OER platforms include capacity for faculty to rate and review objects.
Interactive objects like quizzes and courseware are also really popular but not well supported in the repository context.

So, in the end a lot of vendor-hosted software isn’t flexible enough for this stuff, and we haven’t built it into Open Source options… yet

10 of 30

Consortia

11 of 30

Consortia Solve Problems

Resource Sharing

Purchasing

Infrastructure

I work for a consortia named PALCI (Pennsylvania Academic Library Consortium, Inc). we have about 70 members across PA, but also NY, NJ, and WV. Like a lot of consortia we formed in response to needs for libraries to coordinate around resource sharing. To that end, we help administer an ILL service, and we also negotiate consortial purchasing of serial and monograph packages.

This isn’t surprising I’m sure. PALCI, like a lot of regional consortia, began as a way to solve problems through collaboration and cost-sharing. Our collaboration and our problem-solving have matured and we are ready to share in solving a lot more of our problems and this is leading us to increasingly think about shared infrastructure. In particular we have started to think about repository support.

12 of 30

Need for a Consortial IR

Solution for small-sized colleges
Sustainable/controllable cost structure
Multi-tenant/individual branding and theming
Collaborative management
Improvable, scalable, and customizable to support new and emerging needs

Repositories are a problem to solve for many of the members of our network. A lot of them are smaller colleges that don’t really have an option that is sized right, in terms of both pricing and ease of use. “BePrexit,” as it’s been called is a big issue. A lot of our smaller schools want to have a service that is hosted and that offers a similar product and features, but which they feel more invested in and more comfortable with and can afford.

As a consortia then, we really need a viable option in order to host services, which means first that it is sustainable and low cost, but also that is has a multi-tenant option for the consortia, that can offer individual branding and theming.

We also need something we can manage with our partners collaboratively.

And finally, we need something that we can improve, scale, and customize to meet emerging needs. Our members, even our state government, have asked us to provide more support and infrastructure to make open resources more available, but without an infrastructure that we can afford and afford to maintain, we are stuck.

13 of 30

PALs (PALCI/PALNI) working together

HykuDirect Grant

Partner with DPLA, Stanford, DuraSpace (2017)

IR Task Force

Produced Vision Document and White paper

CC-IR (Collaborative Consortial Institutional Repository) Phase 1 Pilot

2018-2019
Stand-up Hyku and test basic features/functionality for consortial administration
Work with Notch8 to provide development and pilot hosting

As I mentioned I work for PALCI. In 2017 we joined the HykuDirect grant project to test out the use of Hyku as a consortial IR hosted by DuraSpace.

We maintain close ties with a lot of other consortia in the US and beyond, and we found that our sister organization PALNI, the Private Academic Library Network of Indiana (it’s a coincidence!) has also been heavily investigating IR services at the repository level since 2012 and developed a white paper outlining a vision for consortially managed IR services.

So our two organizations got together Brady Bunch style to see if we could magnify that consortial problem-solving to figure how to really make an Collaborative Consortial Institutional Repository work.

So in 2018 we worked with Notch 8 to stand up Hyku and test out the basic features and functionality and identify what the gaps were for us.

14 of 30

15 of 30

16 of 30

Hyku

17 of 30

Hyku to the resku?

IMLS funded, 18-month grant to build a consortia-managed Hyku implementation as “Phase 2”:

�“Scaling Up a Collaborative Consortial Institutional Repository”

Multi-tenant installation and management

Roles/permissions
Shared workflows
Theme/brand templates

Cross-institution searching
DOI Services
Two new worktypes

ETD & OER

Business model

18 of 30

Why Hyku?

19 of 30

Hyku Worktypes

Files + Metadata + Relationships(workflows)

For those of you who are more familiar with Hyrax than Hyku, one of the defining features of Hyku, is that it comes with a pre-defined set of worktypes. Each of these worktypes defines the handling of files, has a specific metadata template, and has relationships to collections and admin sets (which can in turn determine workflows).

I was trying to think of a metaphor for the worktype, because I like metaphors, and because I hate blank slides, and I think of it like this:

The idea of a work in hyrax and Hyku contains a couple of different components. The files are like the thing, the object. The metadata is the kind of wrapping paper I’m using, which says whether or not it’s a birthday present or a christmas present or whatever. And the gift tag explains where it belongs and how it got there. And in the end, I have a nice present.

With Hyku you have a preselected set of choices. So I might have a square box type like this, but if I have something that won’t fit in that, I have to set up a new template with a new slate of choices. On the other hand, I have to create a set of work types that have to work for all of my tenants. So everyone has to fit their content into whatever set of box, paper, and tag we give them -- like getting your present gift wrapped at the mall instead of doing it yourself, right. So it is worth our while to get it right.

The first thing we are working on in this project (for reasons related to timing with another state-funded grant) is the OER worktype (bringing all of our title words together now)... So I want to walk you through what our work has been in this area

20 of 30

OER Worktype Development

Environmental Scan

OER Commons
Open Textbook Network
SUNY OASIS & Open Textbooks

Worktype basics

Review with state-based groups of OER experts
Review from external experts

Development of state project to give grants for faculty OER adoption and creation

In order to get it right for the largest number of users, we spent some time on research into what a generic OER worktype should be.

We started with an environmental scan. In the OER world there are currently several large aggregator projects. These bring together resources from across the web in one location for searching. They don’t host content, or they only host some of the content. Most of these sites are fairly small and so these sites usually had really rich metadata. I did a lot of talking with people at these sites about how they gathered (and actually often created) this metadata.

Next based on the features we thought were best in these sites, we created a description of a Worktype for discussion with OER experts in our states as well as outside of our groups, particularly folks at these aggregators. We really wanted to make sure that we would have metadata that would easily export for sharing with these sites.

Finally, we had some concurrent efforts in Pennsylvania to develop some state supported grants to fund faculty adoption and creation of OER. The. Actual funding for this is still in development, but the process of figuring out what infrastructure we might need for a state-funded repository workflow for publishing influenced our choices.

21 of 30

OER Worktype: Files

Activity/lab

Assessments (tests, quizzes, etc.)

Case studies

Diagrams/illustrations/images

Homework/assignment

Lecture

Lecture notes

Student guide

Supplemental audio/video

Syllabus

Textbook

Tutorials

Audio

Still Image

Moving Image

Video

Text

...or a combination of these

>> OER worktype will support multiple digital file types with some specific features

22 of 30

OER Worktype: Metadata

Metadata Application Profile influenced by OER Commons (https://www.oercommons.org/)

Descriptive metadata primarily from Qualified Dublin Core
Specific controlled vocabularies enforced
Education specific metadata from LRMI/Schema.org
Few locally defined properties

23 of 30

OER Worktype: Metadata

Accessibility is a key factor

Accessible OER are used more and impact student success more

>> The OER worktype will include metadata about accessibility features

24 of 30

OER Worktype: Metadata

Highly inter-related material

A quiz specific to a textbook chapter
An adaptation of a resource into another language
A student guide related to a course syllabus
A supplemental video related to an assignment
…

The availability of ancillary materials play a large part in faculty decisions to use OER

>> The OER Worktype will have enhanced metadata relationships between objects.

25 of 30

OER Worktype: Relationships/Workflow

Typical Proxy workflows already in Hyku

Faculty member has a work to submit…

Scenario 1

Faculty member creates account and �uploads
Faculty member assigns metadata

Scenario 2

Library uploads material as a proxy

Scenario 3

Faculty member creates account and uploads
Faculty member creates some initial metadata
Library assigns/updates metadata

26 of 30

OER Worktype?

Informal “peer review” continues after publication

Ratings, reviews, and course adoptions added to content through user interface
Highly influential in faculty decision making

>> The OER collection will have front-end features to support rating and review. It will also feature links between related items.

Unfortunately, not everything we wanted to do just fit into the “worktype” paradigm, some things have to be wrapped individually, so to speak. So we are also going to be doing some work with the collection and interface later in the project.

One of the really unique needs we identified for OER was that the informal peer review process continues after publication through reviews and ratings. Many other OER aggregator and hosting sites provide the opportunity for faculty to review and rate materials and this is considered highly influential in faculty decisions to adopt works.

At first I thought of this as a strictly worktype concern: OER objects should be able to be reviewed after the fact.

But as we talked about it, we realized this is more of a collection-level behavior, and should be handled through the interface and stored in a separate database from the core metadata. We aren’t working on this aspect of the project yet. We are going to move it to the phase where we do more interface-related work. At that time though, we will also explore how to combine the rating and review data with the object-level metadata for export. These ratings get included in downstream aggregated sites such as SUNY’s OASIS OER search.

I mentioned earlier that we would be modeling a lot of relationships within objects, and we are also hoping that that will be usable in the interface as well.

27 of 30

Project schedule

OER worktype: in progress!

Design and Development of other features: Sep 19 - July 20

Multi-tenant workflows
Individual theming and branding

Pilot: July 20 - Sep 20

Business and Collaboration model development: Sep 20 - Nov 20

Closeout: Dec 20.

28 of 30

Plea for more collaboration

Open Source and OER communities would make �great partners

Commitment to and understanding of “open”

Same puppy/beer metaphors!

Desire to change the playing field for libraries and �education
Complementary skill sets
Opportunity to have a huge impact
Technology or advocacy/education won’t solve problems in isolation

29 of 30

Putting my money where my mouth is...

30 of 30

Thank you!

Jill Morris,

Executive Director, � PALCI

Project Director

Gretchen Gueguen

Digital Projects � Manager, PALCI� Project Manager

Kirsten Leonard,

Executive Director, PALNI

Project Director

Amanda Hurford,

Scholarly Communications Director, PALNI

Product Owner