Fig. 6.

Results for the coarser MS partitions of the document similarity graph into: a 7 communities and b 3 communities, showing in each case their correspondence to the external hand-coded categories. Some of the MS communities with strong medical content (e.g., labour ward, radiotherapy, pressure ulcer) remain separate in our content-driven, unsupervised clustering and are not integrated with other procedural records due to their semantic distinctiveness even to this coarsest level of clustering