The Social Graph. It Isn't

If I could assign one piece of required reading this week, it would be Pinboard founder Maciej Ceglowski's provocative blog post "The Social Graph Is Neither."

It's partially a response to a recent Forbes article that likens the "social graph" to crude oil -- as in "Drill, baby, drill" when it comes to the perceived value of the personal data that we are sharing on social networks. But Ceglowski's post doesn't take the typical approach to questioning these sorts of data mining practices and opportunities (particularly as the "social graph" may be most synonymous with Facebook and its perennial problematic privacy issues). In other words, it's not a call for better privacy or better transparency.

As the headline indicates, what Ceglowski argues is the social graph is neither a graph, nor is it social. He contends that today's social networks have failed to capture the complexities and intricacies of our social relationships (there's no graph) and have become something that's at best contrived and at worst icky (actually, that's not the "worst," but it's the adjective Ceglowski uses).

He writes:

Imagine the U.S. Census as conducted by direct marketers � that's the social graph. Social networks exist to sell you crap. The icky feeling you get when your friend starts to talk to you about Amway or when you spot someone passing out business cards at a birthday party, is the entire driving force behind a site like Facebook. Because their collection methods are kind of primitive, these sites have to coax you into doing as much of your social interaction as possible while logged in, so they can see it.

But if today's social networks are troublesome, they're also doomed, Ceglowski contends, much as the CompuServes and the Prodigys of an earlier era were undone. It's not so much a question of their being out-innovated, but rather they were out-democratized. As the global network spread, that mass marketing has given way to grassroots efforts.

"My hope," Ceglowski writes, "is that whatever replaces Facebook and Google+ will look equally inevitable and that our kids will think we were complete rubes for ever having thrown a sheep or clicked a +1 button. It's just a matter of waiting things out and leaving ourselves enough freedom to find some interesting, organic, and human ways to bring our social lives online."

Whither the Education Graph?

It's hardly surprising, with the buzz around the "social graph," that education companies have started talking about their building an "education graph." I first heard this phrase -- or at least the phrase "the student graph" -- from the textbook rental company Chegg earlier this year, following its acquisition of several other edu startups and with its plans for deeper integration with Facebook. Indeed, when Facebook expanded its Social Graph API this fall, I wrote an article about the possibilities for education startups to tap into it.

But in light of Ceglowski's article, should we ask the same set of questions about the education or student graph? Is it a graph? Does it capture education? Does it represent "the student" and her or his connection to learning? What does this graph entail and how does it differ from a would-be "learning graph"?

I'd argue that we can probably quite easily transfer Ceglowski's pointed insights about "the graph" from the social graph to the education graph. In other words, we do not have the technical capabilities to map or model this. Education changes over time -- different teachers, different schools, different curriculum, different courses -- even different days of the week and moods that the student (and/or teacher) might be in, amount of sleep they got the night before, meals they ate, crises at home, and so on. How do we graph "teacher," for example, in a way that differentiates between "the person who taught my class" and "the person that I learned from" (they can be different)? How do we graph "math" when I have such different relationships to algebra, geometry, trigonometry and calculus?

Can we build, to paraphrase Ceglowski,"a canonical representation of educational relationships"?

Perhaps we could eventually fine-tune a graph that accounts for courses, readings, subject preferences, aptitudes, teachers, majors, grades, test scores -- a student's educational history. But would this be a graph that's actually about learning? After all, that's what matters, right?

Perhaps such a learning graph will come. Perhaps it will be democratic and distributed. Perhaps. Again, to paraphrase Ceglowski, "It's just a matter of waiting things out and leaving ourselves enough freedom to find some interesting, organic, and human ways to bring our social learning lives online."

I'd love to hear folks' thoughts on this, particularly with the amount of attention that's being paid to the promise of learning analytics.