read8 min
Transparent AI
AI and bias: the importance of critically questioning machine learning technologies
See experiment
Created by Oscar Senar and Gert Frankei
How do the colors used in online news article images shape our perception of the world, and what insights can be gained through large-scale color analysis?
Background
We researched how online news articles paint our worldview in daily life. We collected and analyzed 553.527 images from English articles, highlighted by news.google.com, from July 1st, 2021 until June 30th, 2022 and extracted 2.715.868 unique colors.
What did we learn
Analyzing over 500k images allowed us to work around design principles (communication, color spaces...) from a technical perspective using color analysis JavaScript libraries and data analysis and visualization in Python. At the same time, we faced the challenge of storing, analyzing, and processing such a large dataset in a relational database.
Ambitions
Introducing this research allows us to reflect how we perceive visual information and become more attune to how the media literally paints our worldview. In turn, we become more aware and appreciative of the colors that surround us everyday.
Read, see, play next
see
Understanding environments through atmospheric sensors
Smart watch visualization
read6 min
Visualizing climate mobility in Africa
How different user goals shaped our technical approach