My Digital History at Western server is now up, so I can host research tools. The first is an alpha version of Frobisher, a prototype that mines information from the online Dictionary of Canadian Biography to refine Google searches. The user starts by entering the name of a person. If there are multiple possibilities (e.g., MacDonald), he or she is asked to clarify which person is of interest. Since the coverage of the DCB is limited to people who died before 1930, the system performs best with pre-20th-century queries. This is very much still research software: it does quite well on some things (e.g., distinguishing Sir Alexander Mackenzie from Alexander Mackenzie) and not so well on others (e.g., turning up useful material about William Connolly). There is more information about the design of Frobisher in a series of posts on this blog. I welcome any and all feedback at firstname.lastname@example.org.
Tags: data mining | dictionary of canadian biography | hacking | search | text mining