Thread by @sentiment_io

In Pillars of Tomorrow, we used emotion analysis to extract data through the Twitter API, analyze people’s emotions, and apply them as generative music traits.

A thread on data analysis 🧵👇

In our previous collection, Pride Month Tribute, we collected 100K online articles and transformed the emotions into colors.

The stronger the emotional expression was during a specific time period, the more prominent the color was in the artwork.

This is a prime example of happiness being the strongest emotion as you can see by the dominant presence of yellow which represents the feeling of happiness in our first NFT collection.

For this collection, since the NFT community is most active on Twitter, we used Twitter’s API as the source of our data.

The Twitter API provides free access (essential access level) to anyone interested in collecting and analyzing data about users and tweets.

According to the (GDPR), one can retrieve tweets published in the last 7 days, without exceeding 500,000 tweets per month.

Data and results for our upcoming music NFT collection in creative collaboration with @_dahleea_ 👇

Our data is derived from emotional analysis models to extract the emotion scores (happiness, anger, sadness, fear) from approximately 105,762 tweets, written in English, and containing mention of ‘Tomorrowland' during the time period between July 15, 2022, and August 12, 2022.

The analysis showed the following results:

Happiness: 38%
Anger: 8%
Sadness: 26%
Fear: 28%

On an additional note, here's a very good thread if you want to learn more about Machine Learning models by our friend @bubits_

https://twitter.com/bubits_/status/1529444488455897094

Stay tuned for the next thread on how these emotions are transformed into generative music traits.

Emotions are key to our generative artworks and we are determined to design innovative ways to showcase people’s feelings.

#CodingIsArt