In physics, the generators of the fundamental representation of $SU(2)$ are taken to be $\sigma_i/2$ instead of $\sigma_i$ where $\sigma_i$ denotes the $i$th Pauli matrix. This is because, as I understand, if we take $$T_i=\sigma_i/2,$$ the Lie algebra of $SU(2)$ looks identical to that of $SO(3)$. But why would we like to match the $SU(2)$ Lie algebra with that of $SO(3)$ in the first place?
Or is it the other way around? Is it because the $SU(2)$ Lie algebra is identical to the $SO(3)$ Lie algebra, we make the choice that $T_i=\sigma_i/2$? But in that case, I don't know how to derive the $SU(2)$ Lie algebra without starting from any matrix representation of $T_i$'s.