 One of the research subgroups in DCC is the subject analysis group led by Dr. Miles Efron, which aims to increase coherence and comprehensibility among the diverse content of DCC. In our recent work, we have been working to map out major topics among the range of content in DCC and integrate them in a meaningful way into the interface. Using a technique called latent Dirichlet allocation, we seek out areas of commonality amongst the content. However, with the scale and inconsistencies among metadata sourced from institutions across the country, one very quickly runs into problems with messy records, ones with little descriptive metadata or overwhelming amounts of administrative metadata. Thus we identify overly similar items and ignore them while mapping out topic areas. The difference is quite striking. We are currently testing ways to make these topics meaningful to online users. First, we name topics based on representative terms. When you look at an item in DCC, you're shown topics about that item and related items within those topics. For example, in this screenshot from our test bed site, the record shows items from the Machine Burrows Institute topic. We are also working on central topic pages where you can explore these groupings. If you're interested in our work, please contact us.