The idea of the Department of Happiness is based on idea of getting feedback from users based on weighted descriptors.
My habit tracker has a version in the mood collection interface. First, the user is presented with a list of descriptors they've compiled:
Once they choose a descriptor they can give it a weight:
These descriptors will vary over time:
I wanted then to collect as complete a digital profile as I could of the user. Then, use both historical trends and patterns from other users to try and extrapolate decisions that would probably increase their self-reported happiness.
The main focus was setting up the population of events. The computer is trying to group people together and expose them to both situations and substances.
The computer might recommend a group for a festival and suggest that their experience would be more enjoyable with weed.
While and after the duration of the event, the participants can give their opinions in the form of weighted tags.
The band might be Melodic
59% and Energetic
77%. Group member Bob is Cute
60% and Intereating
-35%. The user self-reports as Entertained
40% and High
40%.
The app could prompt the user for particularly salient descriptors.
From this information the computer builds a model of probable responses to various stimuli.