There are several ways to do so (simple Koopmans' theorem approach):

First, "optical criteria": broaden your calculated DOS with a Lorentzian or Gaussian function [in my old setup, 0.75eV was a good FWHM]. The difference between calculation and experimente is usually bigger than the differnece of the functions, so take what you like, just declare it correctly. If you do XPS of molecules and you have isolated orbital energies, you may as well just use "Delta functions" at the orbital energies.

Then, actual matching: you should calculate a reference for which you know what you have got, e.g. your bulk substrate, your untreated molecule or something like that. Then you can align orbital/band energies to a certain binding energy value. With this alignment, you are now able to compare the relative orbital/band energies of your postulated structures to your measured data to achieve an assignment.

Of course, this is all the most simplistic approach, for adsorption on metal surfaces you may have additional effects that need to be covered by more augmented calculations, e.g. using the final state approximation, but I personally have never needed that so I can't help with that.