An important question. At the OER Foundation we have found that the internet is the best repository providing the agility for a wide range of ‘decentralised’ applications where the sum is greater than the parts. Diversity builds resilience into the ecosystem.
I agree with the approach of “put your content in a lot of different places.” Our IR is harvested by major search engines. All of our resources are submitted (when they qualify) to the Open Textbook Library, OER Commons, Merlot.
Digital preservation, stable handles/avoidance of broken links etc. and version are good reasons to look at repositories. There are limitations with any system, so thinking through your use cases is helpful. Here is what we do at Virginia Tech. We run DSpace in house as our IR. Here is our list of Open(ish) Textbooks https://vtechworks.lib.vt.edu/handle/10919/70959 (yes, one has an ND license and some are not textbooks). I upload as many file types as are available – including sourcefiles. (We use Pressbooks for publishing and use other software for production when its better fits the needs of the book.)
I created this two-hour video recording of demos by various platforms and PDF Matrix in 2017 which may be of interest: Seven Platforms You Should Know About: Share, Find, Author, or Adapt Creative Commons-Licensed Resources This is prior to LibreTexts and before OERCommons made significant improvements to their OpenAuthor so it is not exactly up to date. It has a U.S. and American english bias and doesn’t include software that are outside of what we could reasonably consider institutionally. (It is unlikely that I could get permission to spin up or pay for hosting of an instance of Moodle as we already pay for a different hosted LMS/VLE.) So, that are some gaps here but I hope you find this useful.
I am not the repository manager but can put you in touch with him if that is of interest.
Yes, OERTX is a (as you know, award winning!) microsite powered by OER Commons. Microsites are ISKME’s comprehensive solution for hosting and indexing a unique content library while facilitating collaboration for a specific targeted audience. Microsites are best suited to organizations who seek their own branded library with its own domain that sits apart from OER Commons.
You know I more than agree with you Wayne. But where on the internet would an institution suggest putting stuff? I worry some seeing a lot of shared items that are Google Docs/Drive files, with their gibberish web addresses, and do not seem very durable.
And do we count on keyword search to find? any fans of taxonomies? Asking for a friend
Thanks for sharing this Anita (and cheers from us as it was an Open Education Week 2017 resource).
I also like the range and quality of the BCcampus Open Directory
But how to we avoid the loss of things such as the OER World Map?
Again - a good question. I think that the ‘internet as repository’ helps us to think about a shift in emphasis from tools (technology) to educating educators (pedagogy) about what is important when deciding where to store and share OER. Perhaps its useful to think about the criteria educators should consider when deciding on spaces to put stuff, for example:
- Open standards
- Open file formats
- Free and open source (to mitigate against vendor lock-in, and encourage others to adopt, adapt, and make available, i.e. resilient decentralised nodes, on same terms, with same FOSS tech)
- Search functionality including search by license type etc.
- Discoverability (indexable by external search engines; semantic construction for better parse-ability; clear, thoughtful ontological hierarchy and tagging; clean, readable URIs to allow for easy referencing and sharing of links; open licensing to allow referencing and reuse)
- Versioning and forkability!
- Attribution builders
- others? (PS - made this a wiki post so others can add to the list
- Digital skills development - invest in digital literacy skills, particularly copyright literacy, for staff and students so they understand how to create and curate open content ~ Lorna
- api’s (Clint)
- A highly configurable and customizable UI (and hiring a UI/UX expert to conduct user testing) Pay attention to what your human users need (which is a UI that works for them) and what the machines (Google, OERSI, etc) need (aka education resource specific metadata like LRMI) (Clint)
Nice one @mackiwg - I’ve expanded a couple of those thanks to the wiki feature.
Me too, we love wikis!
I’m a proponent of using OER in an LMS, preferably an open source one, which leads me to promoting both Moodle and MoodleNet.
Moodle is free and opensource. MoodleNet is free and opensource. The more people that use them the greater the support network. An institutional instance is not necessary; an instance for one class is economical and doesn’t require anybody from IT to be involved. MoodleNet will handle most kinds of file formats. I think you could upload a D2L course file if you wanted to (I haven’t tested that.) Moodle courses can be constructed so that all activities and functions work in Canvas, D2L, and Schoology. I have tested all of those. That means that any of the 27 math courses that are Moodle files that I’ve uploaded to MoodleNet will work in any of the other LMSs as well. Search on GeoGebra at MoodleNet
I think we need to expand our thinking as to what a ‘book’ is. Higher Ed is still enamored with the notion that a book is something that must be capable of being printed on paper and that when printed it should look the way that books have been printed for the last 5 centuries or so.
I think a book that is used in classrooms of any kind needs to have activities for students to do that foster learning. The book should have a way for students to talk to each other and the teacher built in. The book should have a way for the teacher to present all manner of assessments to students in many different ways. I’m not a big fan of analytics but some people think it’s important. Interfacing with AI is also something that a book might beneficially do.
Here’s the presentation I did for OpenEd22 which is focused on elementary and secondary (K12) and those classrooms that don’t have reliable or any access to the internet.
At the University of Edinburgh we’ve also gone down the “web as repository” route. Rather than having a single centralised OER repository, we encourage colleagues to share resources wherever they can be most usefully managed and found. Our OER Policy states:
The University recommends that open educational resources should be published
in an appropriate repository or public-access website in order to maximise their
discovery and use by others.
The university maintains a number of different repositories and services for different resource types, e.g. Media Hopper Create for media assets, Open Media Bank for MOOC videos, Data Share for open data, Edinburgh Diamond for open ebooks and journals, Luna for Image Collections. We also share open resources on commercial platforms, including OER for school teachers on TES Resources, 3D models on SketchFab, images on Flickr. Our academics have also created open content on Github pages and we have an old OER Commons account that we’re planning on reviving. We don’t have a single aggregation of all the open licensed content shared across all these channels, however the OER Service maintains a showcase of selected OER on Open.Ed.
There are downsides to this approach of course; we never have oversight of all the open content being created across the university, and there is no single entry point to cross search multiple repositories, however we believe this approach is more sustainable and user friendly than having a single OER repository. If we could solve the cross search issue that would be great, but it’s beyond our resource at this point in time.
One of the reasons the university adopted his approach is that we learned from the experiences of the UKOER Programme (2009 - 2012). That programme invested a huge amount of funding into the creation of a national OER repository which proved to be unsustainable and which ultimately had the plug pulled on it resulting in the loss of a significant amount of content. Some content did survive on commercial web platforms though and you can still find OER created by this programme if you search for UKOER on flickr, youtube and twitter. I do find it interesting that there has been a resurgence of interest in centralised OER repositories over the last couple of years. I hope these new initiatives will learn from the experiences of past projects and manage to avoid some of the pitfalls.
Thanks @danmcguire for posting this. MoodleNet is an open source and free OER repository that you can have today for your institution, this version is only a year old so is still building; you can see what’s happening on https://moodle.net.
Reach out if you need help or advice with it!
I should also note similar to @LornaMCampbell 's situation, we have OER in other places too. I think it is really important that content live where it fits best, and where it is findable. I work in a library, so we also value digital preservation – so that we can preserve the historical record for the future, and re-create resources in case of disaster and technological obsolesce.
We have open source virtual reality and web-based virtual animals in GitHub and linked from a page about the virtual animals which resides in our VetMed library, other tools like the pKAnalyzer live in Github, and while we have MP4s of theatrical performances for Storytelling on Screen stored in our DSpace repository, they are also replicated in YouTube. MP3s from Voices of Virginia are likewise stored in VTechWorks but also available on SoundCloud.
We have interactive problems sets for aerospace engineering stored in CanvasCommons (sample problem in CanvasCommons) but also uploaded to our DSpace repository, VTechWorks.
Some of our content is in Pressbooks and no where else.
We have had our work deleted by others. We hosted figures from open textbooks in WikimediaCommons so that they could easily be found and repurposed. We recently moved to use InternetArchive to host original, openly licensed figures as an overeager wikipedia editor disliked the filetype and lack of notability of the figures as well as the fact that we posted them prior to various books being published (because we wanted to cite them in the book.) We don’t really have extra time to negotiate to get these reinstated. That’s not really part of our workflow so these are presently still deleted. (If you’d like to take a try at getting them reinstated, please do so and let me know what happens!)
In general, we try to share openly licensed content as broadly as possible. But, when we don’t have ownership or control over a hosting site there are risks of deletion. So, we always try to ensure that there is a copy preserved that we DO control. Otherwise, we create more work for ourselves and perpetuate the broken links problem neither of which are helpful.
There are a lot of library initiatives to link networked data. Much of this conversation has moved into the Metadata realm under the name “Linked Open Data.” I think that conceptually aligned with the direction this thread is going. This is not my area of expertise but I have colleagues that specialize in this area. For a tutorial on linked open data, see here or see a second example on linked open data from europeana.
This aligns with the vision by @Mackiwg of web as repository – it might be messier, and not as neat as the dreams of one giant repository for everything (the thing that was common on my first years when the focus was on Learning Objects, remember them?).
The alarm of UKOER should be heard also also a shout out to the Internet Archive as many projects/resources i have been part of at least have traces or more there.
And thanks @Arwalz too for the resources on Linked Open Data as that is something I hope to learn more about-- I am signed up for November’s Wikiedu Wikidata Institute and have hopes to organize more OE activity and action there.
The Orange Grove is another repository that disappeared. I asked (maybe the wrong people) if there was a copy of the content stored off line, and sadly the answer was no.
There is an interesting twitter thread which impacts some of this discussion
in the sense that given the way that search engines are increasingly commercialized and lots of people are making a living selling SEO relying on them to make OER discoverable may not be adviseable.
Thanks for sharing - the University of Edinburgh’s approach in using a range of tools for different media assets is impressive. I see that you’re also using a healthy dose of open technologies under the hood to power your repositories (Eg. Kaltura, Drupal). Of interest, does the University host these services in-house?
Totally agree with your wiki contribution above emphasizing the need for digital skills development which ideally should incorporate a wide range of internet technologies and skills.
I’m curious to know what is driving this renewed interest in centralised repositories. For example, is this a sign of maturation in the field with increased adoption of OER or is this indicative of a growing number of “new entrants” discovering OER? Those who have burned their fingers with repositories that have shut down are likely to be less enthusiastic with the central repository concept.
At any rate, for better or worse, I think the interest in repositories is likely to grow. UNESCO’s first consultation on the implementation of the OER Recommendation with member states has survey items asking about the existence of national or institutional OER repositories and/or plans to establish such repositories.
I’m wondering if the timing is right to start thinking about platform cooperatives for OER as a more sustainable approach to provide shared open infrastructure. A community garden approach where institutions work together for common good in providing these kinds of services.
Perhaps this is a viable model for seeding the use of Fediverse applications in post secondary education (for example Pixelfed for images, PeerTube for video.) The power of the fediverse is the ability for any institution to host their own instances of these technologies while still being able to connect with existing instances. The platform cooperative could be a space for people to build knowledge and experience in using these tools without the associated risks of ‘vendor’/‘institution’ - lock in.
Lots to think about.
Thanks a lot for your contributions, sharing experiences and insights.
My guess with regard to the “renewed interest in centralised repositories”, from my experience at the University of Geneva.
For research papers, the university has an open archive, described as digital repository of its “scientific patrimony”.
Similar to the research mission, the teaching mission is coming forth with more openness and institutions are wondering where and how to store this “teaching patrimony”.
At the same time, in research, the SNSF has just made compulsory the use of the DORA CV . Within this new approach, all publications, funding information, award, etc. have to be uploaded on the ORCID plateform and the SNSF CV links directly to that plateform.
If we foresee a similar move for teaching assets shared as OER, this would mean an equivalent to ORCID plateform or meta plateform from which all the OERs could be retrieved.
Two big differences though: scientific papers have a DOI but OERs do not. Scientific papers are not meant to be adapted, etc. where as OERs are. Maybe expert librarians among you have ideas on that and where the debate is heading?
So, “internet as repository” is fine but usually institutions, by law or not, need to have mastery on their “patrimony” which means either a local copy or an insitutional repository.
Thank you for initiating a wiki on this more than important topic: criteria to make choices. Criteria for educators, but also criteria for learners and in the present case, criteria for decision makers. Could you please make the URL of the wiki explicit?
It is funny Alan that you started the thread with a link to the OERSI presentation because this is how this discussion started. Colleagues in the German speaking part of Switzerland, at ZHAW, intend to implement a national OER archive with this technology https://edu-sharing.com/, inspired by the Germans: Open Educational Resources Search Index.
Also in Nantes, Colin has a project and develops an AI powered OER search tool: https://discovery.x5gon.org/
Given all this information, and given the fact that we do not need to find a solution for today but for tomorrow, where will Open Education be tomorrow, i.e. in 10, 20 years time? What are your thoughts? Which criteria for OER repositories / search tools / etc. do we need to consider today to ensure that OE will thrive sustainably and openly tomorrow?
Thank you Barbara and all for this very interesting thread. I didn’t know about x5gon, checked it out and just tried to email their contact addresses but they bounce. Is x5gon still an active project?
Curt Newton, Director MIT OpenCourseWare
Also of possible relevance (and for which I have again not a lot of direct experience), the OER Hub from Open Education Austria shared at an OEGlobal 2022 Learning Lab
This of course is the big question I hope we can collectively chew on. To me, the criteria should be is the content is portable and can be stand alone in standard web format, or at least not depending on a single software or platform (tremors of Flash).
That’s why I am a proponent of H5P…
It’s always good to get positively Downsed!
Thanks @Downes who has ears and eyes everywhere online.