So far, so good

By Eleonora Rossetti

The poster is an overview of the work, it’s an 11×11 matrix and includes 121 days of Finland, from the first of January to the Thirty of April. 


How does a machine perceive our contents? How can it evaluate our feelings? In another world: what’s the state of the art of AI when it comes to sentiment analysis?

“So far so good” is my attempt to explore “art and artificial intelligence” from this point of view. I used to keep a diary since I arrived in Finland: a good metaphor to talk about personal data(set), I thought.

My first 60 days in Finland evaluated by a Sentiment Analysis Machine Learning algorithm. The clock represents both the passing of time and the one used to synchronize digital devices.

I used A simple Machine Learning Sentiment Analysis algorithm written in Javascript to get an evaluation of every single Finnish day in a 0-1 range.
I had to input relevant words to let the program understand, that’s why I’d say I had the possibility to talk WITH the machine while I was trying to talk ABOUT the machine. 
I turned every numeric result into the opacity percentage of a square, creating a grid. The Grid stands for the pixel structure, the “atom” of the digital world. But it also recalls the shape of digital calendars, a metaphor of the passing of time of which my diary try to keep track of. 

0-1 is the range of the opacity in CSS but also the well-known digits of the binary code. The simple visual output wants also to take into consideration topics related to the digital world like data transparency, privacy online and the simplification of our identities performed by the machines.