Revealing and Using Emotion Detection

From The Theme
HUMAN MACHINE INTERACTION AND SENSING

WHAT IF?
What if we could develop systems and algorithms that sense and communicate human emotions accurately and appropriately?

WHAT WE SET OUT TO DO
We set out to investigate fundamental issues surrounding how systems should handle emotion detection, including optimal approaches for revealing emotional information. We explored if and how an interface should reveal that it has detected a user’s emotion. We also explored the effect of feedback, information and recommendations that matched or did not match the detected mood.

First we induced happy emotions in our participants by having them watch happy videos and describe a great day. Next, we assessed four different ways that a system could provide feedback about its detection of a happy person’s emotional state and offer ostensibly personalized media recommendations. Participants received one of four different types of feedback and were offered either happy (emotion-congruent) or sad (emotion-incongruent) media content by the system.

WHAT WE FOUND
Our study affirms that feedback is a design necessity – it helps users understand affective computing systems and produces more positive responses to the systems. Our findings illustrate that affective computing, emotion-based personalization, and similar feedback clearly informs users perceptions and attitudes. These elements affect users opinions on the quality, trustworthiness, and difficulty of a system, as well as their attitudes towards, and memory of, recommended content.

PEOPLE BEHIND THE PROJECT
The Late Cliff Nass was the Thomas M. Storke Professor of Communication at Stanford University and held courtesy appointments in Computer Science, Education, Law, and Sociology. He was also affiliated with the programs in Symbolic Systems and Science, Technology, and Society. Nass consulted on the design of over 250 interactive products and services for leading technology and consumer-electronics companies.

Shailendra Rao has worked on a number of technology projects and companies, including Uber, LIifelock, Google, and Vox Media. Dr. Rao received his Ph.D. in Human Computer Interaction from the Department of Communication at Stanford University in 2010. Dr. Rao was a researcher in the Communication Between Humans & Interactive Media (CHIMe) Lab at Stanford University from 2004-2010.