We wanted to work with analyzing emotions
What it does
Deanna Troi is a Slack integration that monitors the emotional range / mood in a Slack channel and responds appropriately when Slack channels reach certain thresholds of emotion. For example, if anger in a channel becomes dominant, Deanna will post some happy cat pictures, inspirational quotes, or "chill out!" messages in order to diffuse tension in the channel. As female software developers (plus Chris!), we are concerned about the levels of harassment sometimes faced by women who enter the public sphere online. We wanted a way to monitor communications, and, if they have too many trolls or negative comments, to try to redirect the conversation to a more positive path.
How we built it
Python, Flask framework, deployed on Heroku, using the Slack API, using the IBM Watson sentiment analysis in order to rank emotional content.
Challenges we ran into
We did not successfully integrate all the pieces and deploy the app
Accomplishments that we're proud of
Integration of different APIs
What we learned
What's next for deanna
We'd like to get Deanna to have a wider range of responses to different kinds of emotions (right now it just ranks anger, joy, disgust, fear, sadness). And also get Deanna working in multiple Slack channels.