Approach I
Using
quantum mechanics, one can derive the relation between the oscillator
strength calculated from the absorption coefficient and the transition
matrix element |x|2. And the spontaneous decay rate is also decided by the matrix element |x|2, therefore, one can find the decay rate from the absorption coefficient.
f=\frac{2m\omega}{\hbar}|x|^2
while spontaneous decay rate:
A=\frac{e\omega^3}{3\pi\hbar c^3\epsilon_0 n^2}|x|^2
Therefore,
\frac{A}{f}=\frac{\omega^2e^2}{6m\pi c^3\epsilon_0n^2}
Approach II: Judd-Ofelt theory,
f=\frac{2m\omega}{\hbar}S
A=\frac{\omega^3e^2n^2}{\hbar c^3\pi\epsilon_0}S
Therefore,
\frac{A}{f}=\frac{\omega^2e^2n^2}{2mc^3\pi\epsilon_0}
One can see that, in two approaches, there is an factor of 3n^4.
I am not so sure why two are not the same, but at this moment I tend to believe the second approach because it is specialized just for this case.
Usage
In a real experiment, oscillator strength can be found using partial sum rule:The dimensionless oscillator strength is:
f=\frac{2c}{N_e\hbar \pi \omega_p^2}\int_{E_2}^{E_1}n\alpha dE
Here
\omega_p^2=\frac{e^2 \rho}{m\epsilon_0}
[Reference:]
Kumar G.A. et al, Journal of Luminescence vol. 99, p. 141-148 (2002)