utku said:
I think you want to say homogeneous and anisotropic spaces.
Yes, I meant ANisotropic, in both places!

Thanks!
utku said:
Can you express more explicitly the term 3-parameters and 4-parameters groups please?
The number of parameters in a group is the number of variables you need to specify to pick out a particular group element. For example the usual group of 3-d rotations acting on a sphere S - you can rotate the North pole P into any other point Q on the sphere, and you must give the latitude and longitude of Q - that's two parameters. But after doing that, you can still rotate through some angle ψ about Q, keeping Q fixed. That's a total of three parameters. And the group action on S is multiply transitive - since ψ can be anything, there are many group elements that map P into Q.
The space section of a Bianchi cosmology has a 3-parameter symmetry group, because you can map P into any other point Q, and this time Q needs three coordinates to label it - x, y, z for example. And there is no further symmetry, you can't rotate about Q due to the anisotropy. So the action is simply transitive.
The Kantowski-Sachs cosmology has one more parameter - there's a symmetry axis at each point, so you can hold Q fixed and rotate through angle ψ about this axis - making it multiply transitive.