🏢 Global architecture

The GALACTIC project is architected using Python’s capability to detect extensions through the mechanism of entry points. This architecture has allowed us to develop a core that is agnostic of its extensions, offering developers the ability to add their own description mechanisms or exploration strategies.

The core of GALACTIC is divided into several layers onto which extensions can be added:

• characteristics that allow the extraction of any type of value from the data;

• descriptions that describe a set of values (and thus data) through the boundaries of a generalized convex hull;

• strategies that explore a dataset by focusing on groups that satisfy certain monadic predicates;

• measures that feed meta-strategies;

• data readers that enable the reading of any type of file to populate datasets.