Unsupervised stylistic segmentation of poetry with change curves and extrinsic features. Brooke, J., Hammond, A., & Hirst, G. In Proceedings, Workshop on Computational Linguistics for Literature, Montreal, 2012.
abstract   bibtex   
The identification of stylistic inconsistency is a challenging task relevant to a number of genres, including literature. In this work, we carry out stylistic segmentation of a well-known poem, The Waste Land by T.S. Eliot, which is traditionally analyzed in terms of numerous voices which appear throughout the text. Our method, adapted from work in topic segmentation and plagiarism detection, predicts breaks based on a curve of stylistic change which combines information from a diverse set of features, most notably co-occurrence in larger corpora via reduced-dimensionality vectors. We show that this extrinsic information is more useful than (within-text) distributional features. We achieve well above baseline performance on both artificial mixed-style texts and The Waste Land itself.
@InProceedings{	  brooke8,
  author	= {Julian Brooke and Adam Hammond and Graeme Hirst},
  title		= {Unsupervised stylistic segmentation of poetry with change
		  curves and extrinsic features},
  address	= {Montreal},
  booktitle	= {Proceedings, Workshop on Computational Linguistics for
		  Literature},
  year		= {2012},
  download	= {http://ftp.cs.toronto.edu/pub/gh/Brooke-etal-CL+Lit-2012.pdf}
		  ,
  abstract	= {The identification of stylistic inconsistency is a
		  challenging task relevant to a number of genres, including
		  literature. In this work, we carry out stylistic
		  segmentation of a well-known poem, <I>The Waste Land</I> by
		  T.S. Eliot, which is traditionally analyzed in terms of
		  numerous voices which appear throughout the text. Our
		  method, adapted from work in topic segmentation and
		  plagiarism detection, predicts breaks based on a curve of
		  stylistic change which combines information from a diverse
		  set of features, most notably co-occurrence in larger
		  corpora via reduced-dimensionality vectors. We show that
		  this extrinsic information is more useful than
		  (within-text) distributional features. We achieve well
		  above baseline performance on both artificial mixed-style
		  texts and <I>The Waste Land</I> itself.}
}

Downloads: 0