The identification of diffuse interstellar bands (DIBs) remains an open problem since discovery of the first two such features 80 years ago. Any attempt to solve the mystery of the carrier(s) of the DIBs must involve interdisciplinary collaboration between molecular physicists, chemists and astronomers. It is expected that progress in this field will be possible when all known DIBs are divided into families in such a way that only one carrier is responsible for all bands belonging to a given family. In this paper we review some attempts to solve the problem of the origin of the DIBs and we present our latest trials on dividing DIBs into spectroscopic families. We define and discuss the problem of 'noisy correlation' between strengths of different DIBs. Noisy correlation has to be controlled carefully when we look for spectroscopic families of DIBs.