Unbalanced data in SAS Enterprise Miner

Question

I am working on a text mining project related to an Enron email dataset. The goal is to create the best model for predicting the target variable (label). The target is binary, 0 being non-fraud and 1 being fraud. Right now the data has about 420,000 0s and 2000 1s, so all of our models are giving high accuracy. I believe it is seeing the 1 records as error. How would I go about fixing this in Enterprise Miner?

I ran my models without resampling or creating a less unbalanced set. Got high accuracy.