Yes, it's actually pretty straightforward. You acquire the spectrum from a biased sample and recalculate from kinetic energy to binding energy.

Ebind=hv+eUBias-Ekin

Then the WF is simply

WF=hv-Ebind,SEC

in which SEC stands for the secondary electron cutoff, aka the spectral cutoff at high binding energies.

Thank you Sir Jürgen Weippert

I don't think it's pretty straightforward for non-conductive samples like silica. I even doubt that UPS is able to measure the work function of such samples.

Unfortunately, that may be true. I may have confused the translation of "fused" with something else.