 "Big data and the big leap forward")
This is more towards the end of the meeting than the beginning. “In a couple of decades, everything will be big data”, asserts Jose Ferreira, CEO of Knewton.
Since I don’t quite know what big data means and what it’s uses are, I look at him blankly.
Everything, he says, again to stress his point. “Do you know what big data is ?”, he asks. My blank stare is giving me away.
I say that if “everything” is soon going to be big data, I better learn quickly. Sounds like there’s little choice.
He plunges in headlong. “Big data is the mathematics of measuring little bits of something. Once you have enough activity in one place – it could be Google searches, Netflix, Facebook posts, Amazon purchases – you could do the math on it and based on that recommend the next thing.” Whenever you have a certain set of data – whatever the data may be - big data is the math of assessing the objects and recommending the next one.
Something clicks. “So when I buy a book on Amazon and it suggests other books that may interest me or tells me that other buyers of the same book I just bought have purchased these other books, is that big data”, I ask.
More From This Section
It is. And Ferreira’s data is student’s homework and answers to questions. Tens of millions of data is produced on a daily basis when students do coursework and homework (Knewton powered) and this data is now captured and analysed by Knewton – all thanks to the advances in big data and machine learning.
We are at the fag end of our meeting when I have finally understood what Knewton is and does. I have told him about Khan academy and Duolingo – I have met the founders of both recently– and neither had to spend 45 minutes to explain what they did. Their models sounded much simpler although everything they do is “free forever”. To my traditional, archaic brain, that sounds like a failed business model.
They are much simpler to understand, he says. Unlike Khan academy and Duolingo, Knewton in a sense works behind the scenes. Even the students may not know that their material is Knewton-powered or that each action of theirs is being examined minutely.
Students doing homework on a daily basis produce mammoth quantities of data. Much more data than say Google or a Facebook get on a daily basis. “All course work has to be Knewton powered and once this is done, every time a students touches anything, it produces data for us. We know our students better than anyone else in the world”.
Based on this data, Knewton can assess what a student knows, his learning patterns and needs, what he is struggling with and what next he can attempt. This helps them develop the next “perfect piece of content” for the student.
“We can now take all the combined data in the world, find a group who is just like you – their learning patterns, their abilities, the knowledge level, the speed, the IQ level, the effort you put in – and we the find the single perfect piece of content to place before you and your intellectual clones”.
This allows learning to become more adaptive – the advantages of which are obvious. Adaptive learning allows lessons to be tailor made for students – they learn at their own pace, absorb the past learning and move to the next stage only when ready.
Backgrounds, interests, attention spans and IQ levels can distinguish one student’s pace of learning from another. Yet conventional teaching methods force learning speeds to be equated across all students.
“Right now, everybody gets the same textbook for their homework. Now that’s bad for a number of reasons. If you are in a bottom half of the class, it’s too hard for you and you quit. If you are in the top half of the class, it’s too easy for you, you’re bored and you quit. Moreover, it doesn’t contain video or games”, explains Ferreira.
In conventional learning, if there’s something you didn’t follow from two years ago, it won’t be in this year’s textbooks. But today you can pluck a concept you haven’t followed from as many years ago and learn and clear it today.
The use of big data allows Knewton to sift which concepts are clear to the students and which aren’t. “A teacher can be told that the class has failed to understand 3 of the 22 concepts taught to them so he or she can repeat what students have failed to follow”.
Unlike traditional teaching, if there is something a student is unable to understand in science because of a concept he may have failed to understand in math, he can access the content on math to help him understand his science better. You can move across disciplines with ease. Otherwise, to learn a concept that may have been covered a year ago in another subject, you’d have to try and schedule an extra class with someone.
Ferreira was working in US education firm Kaplan when the idea that learning could be tailor made for each student occurred to him first– almost 23 years ago. But back in 1993, when he tried to build a course for Kaplan – the world’s first adaptive learning course – he had to give the students so many questions that they didn’t like the course.
But recent advances – in digital technology and big data – have made it much easier to obtain the answers needed without annoying the students quite so much because technology has caught up.
However, unlike Khan academy and Duolingo, with Knewton everything is not free forever. Knewton typically partners with schools and content providers and charges a fee per student. In some cases and countries – depending on the model being followed in the local market – the fee is paid directly by the student but mostly Knewton is paid by its partners – typically the school or the content provider.
The platform – founded in 2008 – is now available in 21 countries in multiple languages and through its partnerships reaches 10 million students. The 200-people company with five offices (New York, London, Tokyo, Beijing and Sau Paulo) has so far raised US $ 150 million in the last few years and it expects to get profitable before this sum runs out. Investors expect the company to list at some point – although Ferreira says it’s not imminent as of now. Fast company has named it one of the ten most innovative companies in education globally and World Economic Forum has called it a Global Growth Company.
Ferreira is here in India to cement his partnership with Tata ClassEdge. He’d just met Mistry in Mumbai prior to coming to Delhi. The future of learning is finally here in India. We have to now see where it takes us.