the new ellipsometers have built-in software to calculate optical constants (n,k). If your does not have such you have to ask a specialist to solve the so called inverse ellipsometric problem.
You certainly would not use Beer's law to calculate absorption coefficient for reasons explained in the 2nd link. As explained in this reference a good way to go is so-called dispersion analysis which is based on assuming oscillators for your different bands and calculating from there the dielectric function of your material. With help of Abeles method you then calculate, based on Maxwell's equations the spectrum of your film. By comparison you can optimize your oscillator parameters and, finally, not only obtain the optical constants, i.e. the complex dielectric function, but also the parameters of the oscillators and thereby also a very compact representation of your optical constants (see, e.g. 1st link).
Optical absorption coefficients (α) were calculated by correcting the reflection using a formula Exp-αt = T/(1-R), where t denotes the film thickness. The α –value of amorphous semiconductors generally follows a relationship of the form
αhv = (const) (hv – Eg )r
where Eg is the optical band gap (Tauc gap) and r is a constant. Using the model with r=2 proposed by Tauc, Eg was estimated by linearly extrapolating the plot of vs hv and finding the intersection with the abscissa (Tauc Plot).