A research project for evaluating how well the details of the non-static logo of a large organisation can be evolved (within a given base structure) by automatically incorporating like/dislike feedback from the organisation's members.
Non-static means a) the basic logo design describes an entire familiy of logos (not just a single logo) by including some variable parameters and b) the intended use of the logo is that those parameters are not fixed over a long time, but vary frequently (e.g. for each use or once per day).
The research is performed by a combination of a software development with a field study.
The software development uses reinforcement learning on a parameterized representation of all possible logos.
The field study uses the new (in 2023) logo of Freie Universität Berlin as an example, which involves colored rectangular blocks in the background that are variable.
Directories and key files:
doc/logo-fub-story-de.md: Explains the idea to the members of Freie Universität Berlin (FUB).doc/...:- ``:
- ``:
- ``: