A quick post regarding our recent experience with Natural Language Processing. First, Professor Matt Erlin from the German department recently presented early prototype-level results of his work on material culture in the nineteenth-century German novel. We've been fortunate to work with Professor Erlin on his project. A sample of the sort of deliverable we prepared for him can be found here.
Second, we've started looking at natural language processing pipelines and platforms, both for performing transformations (part of speech tagging, spelling normalization, etc.) on texts being prepared by the Spenser project, and for automatically recognizing named entities (people and places) in the court records that will be transcribed and encoded under the IMLS grant just awarded to the university. We're currently focused on GATE, a standard package for such applications, and on MorphAdorner, a package build on top of GATE and especially suited for early modern texts like Spenser's.