Predicting population-level engagement in content using neural similarity and single neuron intracranial recordings in humans


Speaker


Abstract

We estimate the level of engagement viewers have in content using a novel set of methods to measure similarity between neural patterns across individuals, and by influencing their neural activity. We use our engagement estimates to test the ability to regulate the engagement. I will show results from a study that tested the ability to predict engagement in content and to assess the variety of constructs that correlate with engagement and multiple applications in Entrepreneurship, Ad-Skip prediction, Political Debates estimates and more.