Topics from the Richmond Daily Dispatch Corpus

Large digitized text collections are of immense potential value to historians but are notoriously difficult to digest, given the near-impossibility of reading the entirety of their content within a reasonable amount of time. Making sense of such collections on the basis of searching with keywords is usually inadequate because it is often hard to know beforehand what the appropriate keywords ought to be. However, while large corpora present these challenges for close reading, digital techniques promise new avenues for engaging with text through “distant reading ” (Moretti, 2005). On recent interest has been topic modeling, as exemplifie