Notes on technology behind cMOOCs: Show me your aggregation architecture and I’ll show you mine

For those not familiar with MOOCs Dave Cormier & co. have a nice video explaining them here. Dave should know as the term:

MOOC was coined in 2008 by Dave Cormier, Manager of Web Communication and Innovations at the University of Prince Edward Island, and Senior Research Fellow Bryan Alexander of the National Institute for Technology in Liberal Education in response to an open online course designed and lead by George Siemens, associate director, Technology Enhanced Knowledge Research Institute at Athabasca University and Stephen Downes, Senior Researcher at The National Research Council (Canada). The course was called "Connectivism and Connective Knowledge" and was presented to 25 tuition-paying students at the University of Manitoba in addition to 2,300 other students from the general public who took the online class free of charge – from Wikipedia

If MOOCs aren’t already on the Gartner ‘peak of inflated expectations’ I’m sure by the end of this year they will as various people are gambling Where there’s MOOC, there’s brass?. Recently projects like Coursera, Udacity and edX have attracted over $100 million in investment to deliver MOOCs. It’s worth reiterating that not all MOOCs are the same and as George Siemens recently wrote:

Largely lost in the conversation around MOOCs is the different ideology that drives what are currently two broad MOOC offerings: the connectivist MOOCs (cMOOCs?) that I [George Siemens] have been involved with since 2008 (with people like Stephen Downes, Jim Groom, Dave Cormier, Alan Levine, Wendy Drexler, Inge de Waard, Ray Schroeder, David Wiley, Alec Couros, and others) and the well-financed MOOCs by Coursera and edX (xMOOCS?).

George’s post was entitled ‘MOOCs are really a platform’, which got me thinking that given the recent announcement that Coursera is offering it’s platform to other partners, including the University of Edinburgh, this falls into the ‘as a Service’ model as used in cloud computing. So Coursera are offering chosen partners (for a fee) ‘MOOCs as a Service’ (MaaS), or using the distinction from above ‘xMaaS’.

One other consideration is whether xMOOCs are really MOOCs. Martin Weller recently questioned if the recent offerings from Coursera et al. are truly open. So ignoring the xMOOCs for now I thought it would be useful to survey some ‘real’ MOOCs and identify if there were any technological trends and possible future directions. This process has left me the questioning if there is a need for more shared guidance and support on aggregation architectures. [This post is mainly for me to record thoughts and resources but it might be something we (JISC CETIS) take forward. If this is something you’d like please comment or share this post to let us gauge the level of interest]

How the data was collected

A list of MOOCs was taken from the ‘(Very) Partial History of MOOCs’ section of mooc.ca (maintained by Stephen Downes). This was added to with UK based examples funded by JISC/HEA. All 14 courses were then visited and examined for technologies used. The criteria for inclusion were that the course had to have finished or started.

Because of the nature of MOOCs there is a grey line between the technology: chosen by the course team; recommended to students; and used by students either as part of their personal learning environment or used to facilitate smaller study groups. A distinction was attempted to only include technologies chosen/used by the course team.

Data collection was also not entirely conclusive due some of the early MOOCs (CCK08  and CCK09) no longer having a web presence and others like MobiMOOC reusing their existing web presence (the 2011 version of the course has been edited to become the version for the 2012 session).

A Google Spreadsheet with the data and annotations is here. Please contact me for edits/additions.

Results

Frequency of MOOC tech

MOOC tech treemapAbove is the obligatory word cloud of the frequency different technologies were used in the MOOCs surveyed (if you are feeling brave you can explore the data as a treemap on ManyEyes).

Two things are apparent in this data. Firstly, email is often a hidden technology. Services like Google Groups, Twitter and Facebook all allow (and in some cases rely on) email notifications. Secondly, it’s of limited use to know what technologies are being used, the important question is how they are being used.

We can get some insight into this from the diagram below taken from George Siemens “What is the Theory that Underpins Our MOOCs?

Siemens: MOOC Architecture

gRSShopper is an application that allows you to define your own community of RSS feeds, aggregates content from those feeds and organizes it, and helps you integrate that content into your own posts, articles and other content [source]

Because of the connectivist theory underpinning cMOOCs gRRShopper is a key component in aggregating distributed activity. It’s worth noting that only 5 of the 14 MOOCs surveyed used gRSShopper, but there is evidence that other mechanisms are in place to preform similar functions. For example in Digital Storytelling (DS106) this functionality is handled by the FeedWordPress plugin which allows administrators to specify feeder RSS feeds and selectively repost content to a self hosted WordPress blog. In PHONAR and PICBOD, which are photography based courses, students were directed to publish work to Flickr using a common tag to allow aggregation of work.

The general sense of it all – distributed, chaotic, emergent 

The ‘distributed, chaotic, emergent’ comes from a recent presentation by George Siemens for EDUCAUSE talking about cMOOCs. It’s apparent from the survey of MOOC technology that course teams are taking a loosely joined set of tools that they are comfortable with to facilitate a shared experience with the learner. As commented by Downes when writing about gRSShopper “the users are assumed to be outside the system for the most part, inhabiting their own spaces, and not mine”. It’s also apparent that people are taking discipline based approaches using tools aligned to study areas as previously described with PHONAR/PICBOD.

Even with the bespoke nature of MOOCs there are still opportunities to start collectively raiding the parts bin. Given the widespread use of Twitter in MOOCs are there tools/techniques required to aggregate and disseminate the course discussions? Given the wide use of WordPress within education are there opportunities for MOOC specific themes or plugins? With the ability to freely record and stream video from a Google Hangout do we need a wrapper to allow comment collection and annotation?

Aggregation!

It’s just not the technology that sits on top. It’s been fascinating to read the blog posts from JIm Groom as he and his colleagues come to grips with the infrastructural issues of running ds106.us. As commented by Stephen Downes, and as Jim is finding, “aggregation kills websites”. So if it’s the aggregation of content that’s the glue in cMOOCs, perhaps this becomes the area of priority? Perhaps this is the area where JISC CETIS can most useful provide or facilitate guidance?  As Jim recently commented:

No where in the raging discussion around MOOCs is there anyone talking about sharing the infrastructural/architectural work they’ve done freely with others – Jim Groom in Integrating FeedWordPress with BuddyPress

So having loosely started with the pedagogy, looked at some of the technology I’m beginning to think aggregation architecture/infrastructure might be the best place where an organisation like JISC CETIS could help. Do you agree?   

44 thoughts on “Notes on technology behind cMOOCs: Show me your aggregation architecture and I’ll show you mine

  1. Jim Groom

    Awesome post, the idea of Coursera as MaaS is fascinating to me and now part of a post on open architecture I’ve been struggling with, you rock. Another thing about aggregation, I know it can’t scale to hundreds of thousands of aggregated blogs just yet, but we found that FeedWordPress was actually not killing ds106.us, or UMW Blogs for all those years. It was self-referential feeds in WordPress that continually spiked the resources and brought the system down. It is a ridiculous problem, and one WordPress core code should prohibit from happening. I recently wrote a post about this issue, which makes me even more confident that the 515 aggregated blogs is not too intensive for the server we have now. It possibly could scale bigger, but that begs the question of whether we want ds106 too scale. That said we are still trying to make the FeedWordPress BuddyPress integration smoother.

    But at the same time I would hate to discount the technical innovations around the scaling of Coursera and models like it, as Scott McMillan points out here. but fact is none of that is transparent, it is proprietary code, and at its herart it is Bb on a new scale. It’s kinda depressing.

    1. Martin Hawksey

      Post author

      I’ve been doing a bit of aaSing around recently so it neatly fits the general characteristics.

      Looking forward to your post on open architecture. It’s been interesting to start and thinking about the practicalities of aggregating/filtering on scale, which the Coursera’s of this world can conveniently ignore (if there is anything you have to say, you have to say it in their house). Its also helpful to have the argument that this area supports the core pedagogy of go forth and create and we’ll aggregate.

      Udacity and edX look like the most likely to share their tech but with so much investment flying around I’m not holding my breath.

    2. Amy Collier

      Jim and Martin,

      Just catching up on this post (thanks, Martin!) and wanted to offer an idea. Jim said that the tech innovations around scaling of Coursera et al are interesting, but proprietary and therefore inaccessible. But, at Stanford, we just developed an open-source platform that allows for that scaling. It’s called Class2Go: http://class2go.stanford.edu. The code is on Github: http://gist.github.com/Stanford-Online/class2go

      I’d love to see someone take this code and build the aggregation/syndication elements that make courses like ds106 so rich. Take a look and let me know what you think. What’s missing from Class2Go and can we work together to develop it?

  2. Jim Groom

    And to this question:

    So having loosely started with the pedagogy, looked at some of the technology I’m beginning to think aggregation architecture/infrastructure might be the best place where an organisation like JISC CETIS could help. Do you agree?

    Absolutely, and I will freely help with whatever you need in this regard, figuring out how to streamline and optimize aggregation and filtering would be an amazing place for JISC to help. I love Venture Capital free-zones ;)

  3. Alan Levine (@cogdog)

    Am very excited to see your focus on the heart of the cMOOCs. The things that can eb scaled, automated are all things that lack much heart, or much individualism.

    IN trying to catch up with Jim’s blogging, I have some new stuff soon on our ds106/UMW aggregation concept, including syndicating syndication.

    The crux of the difference is that in the xMOOCs, you have to GO to their place, get through the gates, and the locus of activity is in their space.

    In cMOOCs or aggregation based ones, individuals maintain and manage their own activities, and thus the locus is on the individual; the course is a hub, a place to draw that activity together.

    There’s good room for both approaches, though scaling bores me to tears.

    1. Martin Hawksey

      Post author

      “syndicating syndication” sounds intriguing. Makes me think activity streams … learning registry … (Google Apps Script, ScriptDb, Google Forms …) … hmmm

  4. Pingback:

    Notes on technology behind cMOOCs: Show me your aggregation architecture and I’ll show you mine JISC CETIS MASHe | Massively MOOC | Scoop.it

  5. Pingback:

    Notes on technology behind cMOOCs: Show me your aggregation architecture and I’ll show you mine JISC CETIS MASHe | OER Tech | Scoop.it

  6. Pingback:

    Notes on technology behind cMOOCs: Show me your aggregation architecture and I’ll show you mine JISC CETIS MASHe | Utbildning på nätet | Scoop.it

  7. Pingback:

    Digital infrastructure for learning materials: update July 2012 : Digital Infrastructure Team

  8. Lorna M. Campbell

    This raises all kinds of interesting issues :) There are very valid questions around the use of the terminology and I’m glad to see distinctions starting to be drawn between different kinds of “MOOCs”. In some instances the “open” aspect is questionable, in others the “massive” and different courses are obviously underpinned by very different pedagogies. It’s important to recognise these variabes otherwise we risk trying to compare apples with oranges.

    I understand your rational for ignoring xMOOCs in the context of this post, but I do think it’ll be important for JISC CETIS to keep an eye on the technology used by systems such as edX and Coursera to see how it develops.

    I also agree that there is some really interesting work to be done in the area of aggregation which I hope we can take forward in the not too distant future!

  9. Frances Bell

    (Comment on spreadsheet) Regarding the CCK08 MOOC – the wiki may be unavailable but the Moodle is still there http://ltc.umanitoba.ca/moodle/#p5802. I was puzzled by some comments I read on Twitter and went to check something I made after CCK08 during CCK09, an animated video using words from Moodle and blogs in CCK08 http://www.youtube.com/watch?v=uilkFoe4hQo

    I realise that you acknowledge this in your post to some extent but I think that there is a real danger in getting bogged down in the MOOC labels and taking a provider-centric rather than a learner-centric approach to research and evaluation of these learning activities (see my comment on this post http://davecormier.com/edblog/2012/07/31/20-questions-and-answers-about-moocs/) Blog posts and other media artefacts may turn out to much more interesting sites and traces of learning than the hierarchical stuff.

    1. Martin Hawksey

      Post author

      Hi Frances – I think you make a very valid point and some of the discussion in the office this morning was around how this area is labelled and the confusion/complication it creates. I agree that there is a very provider centric angle in this post. Hopefully the conclusion around aggregation architecture, whilst still sitting provider side, is about allowing learners to work ‘outside the system’.

      Thanks for the links!

      Martin

  10. Frances Bell

    It’s not meant to be a downer Martin;) – JISC is largely about provision. I am really glad that you can recognise that architectures support contexts in which learners and teachers may or may not do interesting things. The labels issue is a bit worrying though when it becomes about people defending ‘territories’ without acknowledging what has come before, during and after. (Going slightly outside my comfort zone here) I am wondering if soft and hard perspectives on aggregation might yield some better understanding of what learners actually do (as opposed to view/click).

    1. Lorna M. Campbell

      You raise some valid points Francis, and as Martin has said, your comments very much chime with discussions we have had on this subject within CETIS. I wonder if you could explain a little more about what you mean by “soft and hard perspectives on aggregation”?

  11. Pingback:

    #Change11 An update on posts related to MOOCs as at 3 August 12 | Learner Weblog

  12. Pingback:

    Notes on technology behind cMOOCs: Show me your aggregation architecture and I’ll show you mine JISC CETIS MASHe | Connectivism and Networked Learning | Scoop.it

  13. Pingback:

    Syndication Machines (plus syndication of syndication) - CogDogBlog

  14. Pingback:

    The Shapes of Things to Come | TeLS Webletter

  15. Pingback:

    Notes on technology behind cMOOCs: Show me your aggregation architecture and I’ll show you mine JISC CETIS MASHe | Educational Technology in Higher Education | Scoop.it

  16. Pingback:

    Notes on technology behind cMOOCs: Show me your aggregation architecture and I’ll show you mine | Connectivism | Scoop.it

  17. Martin Lugton

    Last week’s MOOCMOOC was my first experience of a connectivist course. I’d expected to reflect a lot on the design of the canvas LMS, but actually it was pretty irrelevant to my experience.

    I definitely agree that we should be thinking about aggregators rather than LMS for cMOOCs.

    The challenge of collating discourse around posted units of content (eg a blog post) is an interesting one. I’d love to see an aggregator that worked with this sort of thing, and collated tweets and comments into a more seamless ecosystem, for example.

    Love the Stephen Downes quote: “the users are assumed to be outside the system for the most part, inhabiting their own spaces, and not mine”

  18. Pingback:

    Any Questions? Filtering a Twitter hashtag community for questions and responses [situational awareness] #CFHE12 JISC CETIS MASHe

  19. Pingback:

    Aggregation rumination « Yaba Yaba

  20. Yishay Mor

    Great post. I’ll tell you what I’d like to see. A tool that allows me to do:

    Nomination: you run a MOOC, I participate. I want to propose a feed for you to aggregate. you can accept the feed automatically, manually, or by public vote (e.g. it has 3 people to speak for it).
    Aggregation: harvest the posts from a large group of feeds and display them in a way that I can easily sift through and find the stuff I’m interested. e.g., let me navigate by tags / keywords / most liked.
    Curation: I want to select items from aggregator brought in, and compose a new item from them. Storify style.
    Correlation: I want to find and display links between items. “Joe says A, Ann says B. But Ali links the two in a surprising way.”

    And – I want it all with zero install. Ideally, as a platform independent embeddable component.

    (from http://www.olds.ac.uk/development-blog/aggregationrumination)

    1. Martin Hawksey

      Post author

      Thanks Yishay!

      It’s an interesting list most of the functionality being found in different web services but not conveniently packaged anywhere as a single service. Would you want to go the whole hog and include a discussion forum, space for assignment administration, webinar tools, …?

      Martin

      1. Yishay Mor

        Is that an offer? I mean, are we discussing this academically or is this something you are actually thinking of playing around with, and we might be able to use for http://www.olds.ac.uk/ in Jaunary?

        To answer your question: no. I would like to keep administration, presentation and aggregation separate. I can use Moodle / google apps (with or without coursebuilder), or wordpress for the first two. I can use elluminate or hangouts on air for webcasts. I think it would not be productive to compete with these platforms.

        Aggregation is another issue.

        Lessig says “The Code Is the Law” http://www.lessig.org/content/standard/0,1902,4165,00.html
        I say “the interface is the pedagogy”. By and large, LMSs embody an instructionist pedagogy. Shared documents / wikis reflect an constructionist pedagogy. We need a technology that encapsulated connectivism, or some variant of. Hence, we need a tool that could be integrated with any old platform, and would allow a mass of learners to “listen” to the voice of the crowd, make connections and use those to construct new knowledge and represent it. In other words, nominate, aggregate, curate and correlate.

        1. Martin Hawksey

          Post author

          I’m at the very beginning of putting a proposal in to develop some cMOOC blocks to assist course delivery. My focus would be ‘finding a way to connect the fragmented pieces in such a way as to provide learners with a sense of knowledge coherence’. Unfortunately for you the start date would be after January but it would be useful to incorporate your requirements as part of my funding requirement evidence.
          Martin

  21. Pingback:

    Visualizing cMOOC data: Extracting and analysing data from FeedWordPress part 1 #ds106 #NodeXL JISC CETIS MASHe

  22. Pingback:

    Coursera praises MOOC-wrapping as they attempt to ban it | Hapgood

  23. Pingback:

    Coursera praises MOOC-wrapping as they attempt to ban it « Mike Caulfield

  24. Jeannie Crowley

    We are using Google+ to help participants aggregate and share content. During week one of the experience, I shared a circle of everyone who had followed the course page on plus (participants had the option of opting out). Since we’re using the Google Course Builder platform, and using casual Hangouts for some activities, plus seemed like a natural fit to encourage sharing easily. The Google Groups and our shared Google Drive are good for collaborative projects, discussions and problem solving but it is very clunky to share content in a Google group or forum (copy the link, navigate to site, start new topic, paste link, post topic). With the built in plus button on many sites (or on your mobile if you have the app installed), members if the circle can share with minimal effort. An added bonus of the circle is participants end up on each other’s chat lists and this has encouraged additional connections and collaboration.
    *Our course is an experiential, problem-based learning environment where faculty experiment with new ways to design online courses to promote learner agency and peer-to-peer collaboration.

  25. Pingback:

    Our favourite posts of 2012

  26. Pingback:

    Sheila’s work blog

  27. Pingback:

    Sheila’s work blog » Ghosts in the machine? #edcmooc

  28. Pingback:

    (M)OOC in a Box: Turning WordPress into an Open Course Reader #ocTEL Jisc CETIS MASHe

  29. Pingback:

    Nothing new under the sun… | Teaching 'E-learning and Digital Cultures'

  30. Pingback:

    Responding to the MOOC hysteria | hblanchett.com

  31. Pingback:

    Notes on technology behind cMOOCs: Show me your...

  32. Pingback:

    Open Architecture: Our Course Could be Your Life | Keep Learning

  33. Pingback:

    Open Architecture: Our Course Could be Your Life | bavatuesdays

  34. Pingback:

    Tapping the rhizomes of open learning with FeedWordPress, (your personal analytical de-cloaking device) MASHe

Comments are closed.