I am building a GLM Model that predicts whether there will be a claim in a certain policy or not. So target variable is either 0 or 1, so my choice of GLM was Binomial, and link function was Logit.

In the data I also have an Exposure column, which has values greater than 0 but less than or equal to 1. 1 means that the policy has experience of 12 months, and 0.5 would mean that the policy has experience of 6 months.

How do i include this information in my GLM? Should I just use offset on exposure, or use log of exposure?

Please suggest.