Sentiment analysis is one of those buzzwords that seems to have so much to offer. Over the last couple years, we’ve begun to see serious applications of mood detection technology. This can range from body sensors to speech processing tools like EI Technologies’ Xpression, a UK-based company that develops speech recognition software founded by Matt Dobson and Duncan Barclay.
Xpression is for patients visiting psychologists for a mental illness. Oftentimes, they are asked to record their emotions throughout the day and report back to their doctor. This task is often forgotten or not done properly, so Xpression is a new app that uses voice tracking to pinpoint a mood. The app reads a user’s emotion by sending short recordings of her or his voice to a remote server; this server then assesses the mood based on factors like loudness, intensity, pitch, and speech rhythm, ultimately it can sense changes in emotion through voice, record these changes, and forward them to the doctor.
Clinical psychologist Adrian Skinner told The New Scientist that:
with conditions like depression, people tend to stop doing things like filling in mood diaries. If this app gives us more complete diaries it could help us better find the day-to-day triggers that raise or lower a patient’s mood.
This brings up the point that in order for such innovations to work, the user experience is kept fairly passive. That’s why after being installed on the user’s phone, the app samples fragments of audio during a user’s calls. It provides output based on 5 key emotional states including calmness, happiness, sadness, anger, or anxiety.
Articulating the long-term vision for the app, a scientific advisor to EI Technologies told Digital Trends, the system “extract[s] acoustic features and let[s] the machine-learning system work it out. Eventually, such analysis will be able to be done within the app itself, removing any need for transmission of snippets of audio.
According to Mobile Marketing, while Xpression was initially targeted at patients needing to reflect on their struggle with depression, the app’s developers have further developed the app to address other illnesses such as Alzheimers, ADHD, epilepsy, and Parkinson’s Disease.