What's the Latest Development?
Research officer Saif Mohammad has created a data mining tool that analyzes texts -- specifically emails, though he's tested it on other types of texts as well -- and calculates their "temperature" according to eight basic emotions including trust, fear, and anger. With it, he's discovered that, for example, men used more "anticipation" words when writing to women, but more "fear" words when writing to other men. Women communicated to other women using more "sadness" words and fewer "trust" words. Next year he plans to roll out a Gmail app that could help users better gauge the emotions behind the emails they send and receive.
What's the Big Idea?
Machine intervention in human communications can provide for more self-reflection as well as improved awareness, says Mohammad: "Emotions are central to our life...There are implications in health, there are implications in social cultural aspects, there are implications in product marketing." In addition to the Gmail project, he is working with the University of Pennsylvania's psychology department to see if his technology could help predict heart attacks in patients based on their written texts.
Photo Credit: Shutterstock.com