Michal Kosinksi visited CMU today and gave a talk entitled “Predicting Personality from Digital Footprints” in which he describing a mechanism he’s developed for providing (sometimes) better than human ratings of personality over a range of common traits (the Big Five).
I found this pretty interesting as it describes a very effective mechanism by which the notion of philosophical grammars could be formed. Additionally, he and his collaborators have created a public instantiation of their tool with both datasets and a live tool which can be used for testing etc.
Hopefully I will be able to use some of what Michal has done as part of the engine in my project.
The abstract provided for the talk is:
Personality traits form a key driver behind people’s behavior, cognitions, motivations, and emotions; therefore, assessing others’ personality is a basic social skill and a crucial element of successful social interactions. However, based on a sample of over a million participants, I show that personality judgments made by computers―and based on generic and pervasive digital footprints (Facebook Likes)―are more accurate than those made by participants’ friends, family members, and even romantic partners. Furthermore, compared with humans, computers achieve higher inter-judge agreement and superior external validity (i.e. are better at predicting life outcomes). In some cases, computer-based personality judgments are even more valid than self-reported personality scores. I conclude by discussing the consequences of computers outpacing humans in this basic social-cognitive skill.