Neither glucose nor cellubiose have natural absorbance. Are you talking about DNS or Anthron assay?
If indeed DNS then take a look at the paper by Abdul Saqib and Whitney, Biomass (2011) and Bioenergy, 35(11):4748–4750 were they compared the response of DNS for different mono- and di-sacarides.
The writers rightfully press the point that one should use a specific calibration curve for each sugar. But if you really can't get cellobiose a factor of [mg/ml glucose equivalent*1.288 ~ mg/ml cellobiose] should give you an estimate of cellobiose. Mind you, this is A PROXY ONLY, calculated using their data, and will probably work only at the range of 2-9 mg/ml. Even worse, this ratio is increasing slowly with the increasing concentrations so use with great caution.
Since specific sugar has a specific calibration curve, the growth pattern of organism changes in the presence of different sugars, so according to my suggestion, you should not convert the glucose absorbance values to cellobiose absorbance .