In recent years, the field of data mining has seen significant works on pattern discovery by output sampling. Very recently, these sampling methods have been applied to sequential data which is of a complex nature. The complexity of these data lies in their structure which has a notorious impact on the speed of computation and in particular on the preprocessing. In this paper, we have shown how to take advantage of the BSP (Bulk Synchronous Parallel) programming model to improve the efficiency of output sampling methods on sequential data.