The last laugh – Microsoft and SWAYAM MOOCs

Microsoft may well have the last laugh in the struggle to build SWAYAM – the Indian government’s flagship initiative on MOOCs. The deal is priced at 38 cr INR or about USD 6 mn for a 3 year period post which the government will handle it. This is supported by changes in regulations which permit colleges and technical institutions to use SWAYAM courses for credit – see the UGC and the AICTE guidelines. It is supported by the NMEICT commitment for re-purposing of NPTEL content for MOOC-based consumption. It is supported by host institutions sharing infrastructure and other support for students taking a partner institution course.

This is indeed a positive development for online learning in India. For the first time, online learning will be an acceptable part of the learning curriculum, formally recognized for their credit power. This may enthuse students and teachers to accept the platform and courses, and give students a way to improve their scores and understanding of the subject.

Of course, this platform is not really as open in the sense that it is not open to all for free or to those outside the education system itself to accumulate credits for future sojourns in the academic system or otherwise. It resides as a component within an existing institutional framework with limitations on use. In its implementation, it is likely going to be in the nature of an elective course (at least that is how I think it will be implemented). Over time, whether these courses actually turn out to be massive, is also a question.

When I helped coin the full name of SWAYAM (Study Webs of Active Learning for Young Aspiring Minds), I had for inspiration Ivan Illich’s famous statement in Deschooling Society (1971):

The current search for new educational funnels must be reversed into the search for their institutional inverse: educational webs which heighten the opportunity for each one to transform each moment of his living into one of learning, sharing, and caring.

MOOC systems are intended to be webs, an institutional inverse and are new education funnels. Just like the Connectivists say that Learning is the process of making connections and that Knowledge is the network, the web-like nature of learning is its most powerful when actualized through technology and digital networks. They are not intended to be closed and institution specific. I hope some of that thinking permeates how the system is conceived of and implemented, else it will be no different from how online learning has happened all these years.

But in the more relaxed sense, even this is a valuable opening move. Perhaps our Open and Distance Learning system can migrate to SWAYAM as well (IGNOU, NIOS, SOUs) and bring with them learners who have by choice embraced the distance learning paradigm.

There are many important considerations for getting this to work for the long term for Microsoft and SWAYAM.

1. Open Software, Cloud and API: Use of open software and non-proprietary cloud platforms and technologies: It is imperative, for long term sustainability, to ensure that software stacks are open and reusable in different scenarios. The system must be interfaced with using Open API based services and must exposes developer SDKs. Tie in to vendor specific platforms or application software on a general scale should be avoided.
2. Design: It is super critical to bring some standardization into MOOC design and development processes while allowing for creativity. MOOCs are/should be designed differently from online learning courses. They become a continuous site of interaction, reflection and knowledge creation rather than episodic learner – course silos.
3. Delivery: Support for MOOC learners (peer and institutional) needs thinking. The host institution may or may not be capable or able to support such myriad course choices. The partner institution, which creates the course, may need to think of how to certify (if there is a manual component involved in assessments).
4. Data modeling and security: This is a really huge piece of the puzzle. While data should be available in anonymous forms to researchers, personally identifiable and behavior/performance data should be protected zealously. Data will have to be modeled too along standards that we have to evolve.
5. Equal Opportunity: A crucial part of this platform’s success will be to allow multiple sources of MOOCs to be hosted on the platform. For example, publishers must be allowed to publish and advertise digital courses found to be at par with the government sponsored content. So too may external organizations, whether HE/SE/FE or not, should be allowed the opportunity to host their courses for fee or free. Otherwise this is akin to creating a government monopoly.
6. Engagement: The single largest determinant of the platform success is going to be engagement of teachers, experts, administrators, students and parents.

Some additional notes from a previous conversation with a colleague:

Firstly, it makes sense to have many platforms if and only if we agree that a common API can be created by SWAYAM that saves everyone time in development and centralizes data. This common API can be loosely coupled with many content repositories. However a core part of the implementation of these APIs by any provider should be that they “talk” with centralized servers for taxonomies (curricular definitions), learner profile data, learning experience data, content and so on and so forth. This could be a middle of the road approach which shall also allow distributed centres of innovation. Do look at the Clever API way of doing this as an example (https://clever.com/). They centralize student information from 30,000 schools and then make them login to a single platform with hundreds of tools and resources – this saves time and brings forth continuous innovations in content, curricula and edTech.

The second part is that the money we are spending will yield very low return if content assets are not leveraged through a proper Content Management and Publishing platform which stores content in raw formats and is able to repurpose and publish to multiple platforms and devices. We are going to save atleast 30% in costs of new development, 100% of the cost in repurposing (or close to that) and countless hours of effort and money in publishing cycles and deployments. Plus we will enable an entire generation of teachers and experts (and even students) to contribute content pieces on a mass scale.

Thirdly our strategy for compute and storage should be to enable the fabric upon which all systems work – so rather than providing a scaled up portal, if we provide enough power to serve applications, services, content and data to downstream MOOC or online/blended learning environments and store learner & teacher experience data and performance on the shared cloud, we shall end up truly leveraging the massive scale that we have. Just as an example, let us assume that for the same topic in an engineering course, all institutions with lakhs of students taking the course, we are able to amass and match student profiles with course performance data (content viewed, apps used, forum activity, test results etc.), then we will have an unprecedented scale to build adaptive learning algorithms and recommender systems. Plus we shall be able to, on a mass scale, exchange taxonomies with available international content repositories in a meaningful manner. This also sets the ground for continuous improvements.

The fourth is our ability to take this to low cost devices, phones and even standalone centers with little or no connectivity. It should be possible to use our CMS and build delivery mechanisms to sync data and content between the remote center and the central computing resources.

The fifth is our ability to build a community that can create and localize content, and evolve to support each other. No government can physically build a national community that is to be so large and connected. Using central services APIs, we can soft-connect every individual in the learning system and allow distributed sites of development.

The sixth is that it gives the chance to practice good governance, since all activity can be monitored/reviewed/analysed centrally alleviating the pains that exist today in manual data collection and analytics. One has only to see the Sathyam committee report to understand the scale of the problems we face in educational data mining in India.