board priority: Project disclaimer*

feb 10th, 2025

* please double click on the "Expand" button.

To you, dearest reader,
Know that something else also dear to me is the maintenance and dissemination of a hacker culture which prioritises conscientious software production above all.

The core of this project is a machine learning model that classifies graphs, mathematical objects consisting of nodes and links between them (so called "edges") that can serve to model things like discretised 3D meshes, molecules, social networks and so on. Graph machine learning enjoys some appreciably noble applications in bioinformatics (e.g. protein classification) and social network analysis (e.g. bot detection on twitter lol), just as it is otherwise used for yucky activities (e.g. invasive recommendation systems). A general solution to these kinds of ethical headaches is to initially release the software under a soft proprietary license like the Creative Commons Non-Commercial (CC-NC) license (essentially limiting the code to be used for research), only to re-release it under a more appropriate license later down the road (maybe copyleft like GPL), which is what I intend to do here (even if the 'ware turns out to be hot garbage).

Datasets:
The bioinformatics datasets (MUTAG, PROTEIN, NCI1, ENZYMES) are all ethically sourced and well-established benchmarks, same thing for the social network datasets (COLLAB, IMDB-B), whose data does not feature private nor sensitive information.

signed,
r.d.p.v.