In the context of DFT, there's a bunch of functionals of the electronic density approximating the exchange-correlation ground-state energy, $E_{xc}[n].$

I wonder what is known about writing exchange-correlation density functionals for excited states.

For example, one could try computing the energies of the first excited states of the jellium in terms of the density and use that, in the context of the Local Density Approximation (LDA), to write an XC functional for excited states. Is this possible? How would one do something like this? Perhaps adding the energies associated to the plasmonic frequencies to the ground state energy?