When it comes to online learning, there are programs galor to assist with math and science. Few, however, address reading comprehension and writing ability.
Shadyside startup LightSide
is building tools to delve into this more ambiguous subject area, using machine learning to instantly assess writing for both academic and business settings.
The CMU spinout and Project Olympus startup is well on its way with the help of a grant for $25,000 from the Gates Foundation. LightSide was among only 29 companies chosen as part of the Literacy Courseware Challenge.
CMU graduates Elijah Mayfield, David Adamson and Carolyn Penstein Rosé first began working on the idea while at school as an open source platform to assist academic researchers and social scientists.
When it comes to assessing writing, machine learning tools work well for some aspects of the process, but not so well for others, explains Mayfield, CEO, and a recipient of a 2011 Siebel Scholarship and named a “Global Shaper” by the World Economic Forum.
Making sure students have the right concepts, are stylistically correct and follow grammatical rules are areas that computers handle easily. Creativity and spontaneity is not as easily addressed.
“It’s not just a magic wand,” Mayfield says. “Software relies on patterns of training examples.”
While a final product is still a year away, LightSide is consulting with individual clients to help them leverage the beneifts of machine learning. Several are heavy weights in the assessment industry, College Board and CTB McGraw Hill.
“It’s all about empowering the student and making sure they know what the strongest and weakest elements of their writing are,” he says.
LightSide employs seven people and is hiring.
Writer: Deb Smit
Source: Elijah Mayfield, LightSide