The rest of project and theory will be added Soon..
The simple prototype: Unsupervised Learning in graph such that our embedding has alot of properties such as preserves orbits(i.e., nodes have the same structural properties) and embedding order(i.e., for all i in [d] for all u,v in V(G) z_u[i] >= z_v[i] iff u is super over v(def.))