Everyone is familiar with the basic emotional labels like happy, sad, angry and surprised. But current psychological science claims that the experience of emotion is constructed from at least three key ingredients:
from our body
Information we collect from
the external world
Mental representations from
Context is king
The same situation can elicit different emotional responses from one person to another, or from one hour to the next. The more you understand the surrounding context, the more accurate and realistic your measurements of emotion can be.
For our system to measure the emotions of a user or group, we build a model of the surrounding context from a wide range of data and features. Depending on what sensors and information are available, the data input may include biometrics (eg. heart rate, breathing, voice, face), content (eg. social media text, emojis, survey results), environment (eg. location, motion, climate, time), and historical data about the user or group’s behaviour. Those inputs are synchronised and processed by our emotion AI engines, which are software packages that sense and measure the emotions of the situation based on scientifically validated psychological models. The system then provides an output of emotional scores and key moments, in real time.
Emotion in Artificial Intelligence
Smarter every day
Technology is finally becoming “smart” in ways that we appreciate in everyday life – helping us find what we’re seeking, organising our plans, providing us with entertainment and assisting us in learning. This is AI, Artificial Intelligence, and it’s everywhere now. But part of what sets us humans apart from computers is our ability to give an emotional context to our thoughts, to feel the mood of the situation so we can make better decisions in the moment. One of the great innovation challenges now is to teach machines to sense our moods: to give them empathy and emotional intelligence.
Currently, AI can tell you the best route to your office from your house. AI can suggest new music you will probably like, or warn you when your bank balance is getting low. But imagine how systems like your car, your mobile, your house or your social media accounts might serve you if they knew what mood you were in. Only if AI is programmed to have some emotional intelligence can it start to make judgements that are optimised for your specific context, at each specific moment. This is why we build emotion AI solutions at Sensum.