When you’re looking for OER for an online course, where do you start?
My guess is that most people would start with a generic search engine like Google or Ecosia. If you’re doing this you’re probably finding you have to wade through a lot of random snippets of content and “articles” and an increasing amount of clickbait and AI-generated nonsense.
Moodle’s humble offering to solve this is an open source system called MoodleNet, which we have been developing slowly over the past few years as a research project. MoodleNet’s main agenda is to provide a way for educators to curate content from around the web (mostly OER content we hope!) in a way that can makes it faster for teachers to find quality content quickly. The idea is that eventually it can be a reliable search engine for OER across the web, Despite the name it’s not just for Moodle users (although it does have extra affordances, such as letting you add resources directly into your courses with a couple clicks)
You can use our main site at moodle.net, or install it yourself and run it locally for you and your friends. Several large installations worldwide do this already.
We recently updated to MoodleNet 4, which brought these improvements:
We now use AI to process new resources and automagically fill in most of the metadata for you, which you can then tweak. We find this improves the quality of metadata around resources, which in turn makes it easier to search for them.
We have integrated a simple gamification system to give you some fun goals while using the site - see if you can claw your way to the top of the leaderboard!
We’ve improved the search page significantly allowing you to zoom in to what you want quicker. You can even save your favourite filters in your profile, so they are used by default next time.
Lots of other UX tweaks.
We’re currently evaluating how much effort we’ll continue putting into MoodleNet, and I’d love your feedback!
Is it promising? Should we keep working on it?
Are there other sites like this that you like better?
This is really promising. I like the AI and gamification features. The UX is also very good. I like how the search results are displayed and the use of colorful tags. I have been looking at the edu-sharing option and it seems that MoodleNet maybe easier to implement on first glance. I would like to see how institutions can set this up and manage it locally.
I have been aware of MoodleNet for some time. It’s great that you intend to continue it, and I hope that resources can continue to be devoted to it. I do believe that a tool like this is a useful feature and (making a competitiveness point for a moment) not one that every world-class VLE/LMS has - none of the regional/local VLE vendors do. If readers will forgive me, here is a quote from our recent report on UK digital K-12 “Trends and Issues of Digital Learning in the United Kingdom” (a chapter in the book https://files.eric.ed.gov/fulltext/ED636595.pdf) :
“One feature familiar to universities who use modern VLEs is the ability to share whole courses and import courses which are openly licensed. Canvas allows all users access to Canvas Commons, its sharing platform; Moodle has something similar, Moodle.net. There is no evidence that course sharing of this sort happens in schools, unlike universities. The key issue that came to the fore in the pandemic was the lack of free (or easily licensable) relevant content.”
Great that you are using AI now to help search for material and populate metadata (always the tedious bit).
For world-wide acceptance it will be important that users report on their experience of using MoodleNet via journal articles, so that others (and governments) can learn from their experiences.
Despite the name it’s not just for Moodle users (although it does have extra affordances, such as letting you add resources directly into your courses with a couple clicks.)
This is an extremely important feature of MoodleNet that almost nobody knows about or understands. Some samples of how to do this and as much advertising as you can afford is needed. They’re not going to come just because you built it - this is not A Field of Dreams.
MoodleNet’s main agenda is to provide a way for educators to curate content from around the web (mostly OER content we hope!) in a way that can makes it faster for teachers to find quality content quickly.
MoodleNet needs a much clearer agenda: I would suggest initially a specific goal.
Start by only accepting OER. Ask contributors to attest to the fact that the content is legally openly licensed.
Create a specific and useful purpose for MoodleNet. Make the priority the ability to share OER LMS courses only. Leave the sharing of bits and pieces to somebody else. Requiring users to put resources into Moodle courses will help trim off at least some of the junk that is already collecting too quickly.
Make it widely understood that Moodle courses will work in other LMSs as long as only certain activity types are used. You will likely need to provide a Moodle course for people to use to create the Moodle courses that are shared and some tutorials about how to easily create Moodle courses using other LMSs (AI?)
The idea is that eventually it can be a reliable search engine for OER across the web.
Let Google or Bing or some other search engine be the search engine for all of the bits and pieces and only-maybe-OER. Stick with what you’re very good at and known for as you move into the very not well understood arena of OER. Most educators and administrators still don’t have much of a clue about OER unless you’re only thinking about Higher Ed.
I’m not sold on gamification. It’s never been high on my priority list when I’ve been trying to get a course built.
Oh, and K-12 teachers don’t write journal articles. Nobody pays them to do it and they sure as hell don’t have time to do it on their own.
Thanks for sharing the new version and asking the big question, Martin. Like you open with, the way we search and find is being eroded, and whether AI helps or clutters remains to be seen (i have been dabbling with Bing’s enabled search and quick plays with exa.ai but still, its a lot of sifting through detritus.
All the features you describe are reasonable (and getting some meta data pre-drafted sounds very interesting) and all the features work as advertised. I’d also question, a bit like Dan might be alluding to, when one “searches” for OER, is it to find whole courses or large units one can use almost as is or the building blocks like individual media or activities. I can’t answer as I am not one who is doing the looking and building of courses.
One thing that may help (if even possible) is setting up such systems so they can also be searched using meta search services like OERSI, Mason OER Metafinder, OASIS (that is me not knowing what it takes). Word is also that the World OER Map is having a comeback.
But hey, we can sit here and opine, but what if we collectively tried and give Martin a boost, and give Moodle Net a test run to add existing content? (exclusing @danmcguire who already has stuff in Moodlenet)?
While we’re 100% focussed on OER (the only licenses you can set in metadata are CC ones), we can not currently 100% guarantee that everything that people add is actually correctly licensed due to lack of a team to do that checking. The two directions there are to
a) leverage AI to do more checking and cross-checking.
b) incentivise the community to check everything.
Thanks Dan, that is definitely a direction we should explore and research with practitioners.
Limiting the format to Moodle courses would be good for clarity as you said in some ways, but it’s also a bit less “open”, in that requires you to be a Moodle user to participate, and they are less reusable in other contexts. You can’t use Moodle formats in anything else, but you can add everything to your Moodle course.
It will also mean a lot less content in the database, meaning much more disappointing search results on your first tries which might mean you never use it again.
One idea I have been thinking about is how to manage content-generation projects where we do a community push on creating whole courses on various topics (and in my mind, these would be focussed on areas critical to SDGs). I could imagine short term funding projects being managed in a way that expert teachers are paid to create and review whole high-quality courses (with content, structure, assessments, everything) that are designed to be shared in systems like MoodleNet where those template courses are a click or two away from teachers facing a blank page.
This e-mail may contain information that is privileged or confidential. If you are not the intended recipient, please delete the e-mail and any attachments and notify immediately.
The search filters look useul, but I’m confused on the MoodleNet search, if say for an earth sciences project I seek things to understand about volcanoes, a search on volcano returns nothing relevant - it seems to be partial matching on “volume”, “voluntad”, “LibreVox” maybe an artifact of not enough resources?
And since I put out a challenge the add resources, I now see I need to post 5 for my items to be visible, it will take some time. The AI detection for a web site’s metadata did a fair first cut, that’s a useful feature… well actually it’s hit and miss. A web site I shared that creates a random persona with photos for creative storytelling somehow got interpreted as “Introduction to physics” so it can wildly be wrong (I do not use the verb “hallucinate” for AI)
The spinning koala “loading” cursor is cute
Some small interface items- the use of a floppy disc icon for a save function is not intuitive at all, I was stumped on my first submission on not finding am obvious “save” button. Also, on creating my profile, my uploaded images and background never are saved, have attempted many times.
This is likely not useful, but it does not feel all that well interface tested. I will try to follow through and get five items in there, but hope it does not cloud thing since not much of my work is or has been assembling course materials.
Congrats on the milestone. Given that “4.x” means UX in MoodleLand, community scrutiny will likely unearth some interesting pieces of insight having to do with discoverability (a core topic in OER work if there is ever one).
Perhaps more important than finding the resource is contextualizing it. MoodleNet eases that process, for sure. It’s also a key part of learning design, much of which often requires deep and broad consideration.
If it’s hard to scale the task of checking licenses, the complex processing needed to adapt resources to diverse contexts is unlikely to scale soon, regardless of AI breakthroughs. (While we may use AI tools when we adopt and adapt resources, much of these workflows would still rely on a deep understanding of learning experiences, something for which AI models don’t get straightforward data.)
