Figure 2: Process (above) and nonprocess (below) sub-sequences identified by the unsupervised algorithm