Curating MIT for Tom Friedman

On my flight to the 2017 CASE Summit in San Francisco, and after several prior articles where we applied our Natural Language Processing to the conference and Tom Friedman's thinking, it made sense to go ahead and see what the engine would actually curate from a partner school directly to Tom Friedman.

We scraped in a few sources of bio data like his Wikipedia page and his own website's bio. The engine, as you'd expect, found dozens on rich insights on Tom like "Pulitzer Prize," "London," and "The New York Times." There were far more nuanced insights, too, like "Kosovo," "Crown Prince Abdullah," "2003 Invasion of Iraq," and much more.

When we loaded that profile in, it created a record for Tom, in a test account, at MIT:

Naturally, were this based also on his ID info from his alma mater Brandeis, system behavior, and more, the profile would be 10 or even 100X more insightful, but nearly 100 rich topical insights is nothing to scoff at; especially when in seconds four MIT news items were curated for Tom:

"Practical Parallelism"

"From the NFL to MIT: The Double Life of John Urschel"

"Romantic and Rational Approaches to Artificial Intelligence"

"What Makes MIT the #1 University in the World?"

These items could wind up on an alumni webpage for Tom, in his mobile alumni app feed, or even in an automatically generated and sent personal digest email from his alma mater.

That last item about MIT being #1 was particularly heavily weighted for Tom given his rich history with academia as a student, professor and visiting lecturer at some of the world's preeminent institutions. "Education" is far heavier weighted for Tom than most people. It speaks to why he's with us here at CASE Summit.

What else might the system soon curate for Tom? The funds he's most likely to give to, events he should attend, and community members he could help or be helped by.

Machine Learning is a truly powerful asset not just towards richly and personally engaging every single alum, but making annual and major gift officers more effective in their asks by telling them exactly what someone is most likely to give to. Come see me, Nick Zeckets, and Ken Keefer, at Summit17.