University of Illinois Urbana-Champaign

Mapping Genre at the Page Level in English-Language Volumes from HathiTrust, 1700-1899

Underwood, Ted; Ballard, Shawn; Black, Michael L.; Capitanu, Boris

Permalink

https://hdl.handle.net/2142/50291

Description

Title
Mapping Genre at the Page Level in English-Language Volumes from HathiTrust, 1700-1899
Author(s)
  • Underwood, Ted
  • Ballard, Shawn
  • Black, Michael L.
  • Capitanu, Boris
Issue Date
2014-07-10
Keyword(s)
machine learning, genre, metadata, classification
Abstract
Using regularized logistic regression and hidden Markov models, we predict genre at the page level in a collection of 469,000 volumes from HathiTrust Digital Library. Accuracy is comparable to human crowdsourcing.
Type of Resource
image
Language
en
Permalink
http://hdl.handle.net/2142/50291

Owning Collections

Ted Underwood PRIMARY