The 'Nishimori line' is a line or hypersurface in the parameter space of sy
stems with quenched disorder, where simple expressions of the averages of p
hysical quantities over the quenched random variables are obtained. It has
been playing an important role in the theoretical studies of the random fru
strated systems since its discovery in around 1980. In this paper, an inter
pretation of the Nishimori line from the viewpoint of statistical informati
on processing is developed. Our main aim is the reconstruction of the whole
theory of the Nishimori line from the viewpoint of Bayesian statistics, or
, almost equivalently, from the viewpoint of the theory of error-correcting
codes. As a byproduct of the interpretation, counterparts of the Nishimori
line in models without gauge invariance are given. We also discussed the i
ssues on the 'finite-temperature decoding' of error-correcting codes and cl
arify the role of gauge invariance in this topic.