A Data-Oriented Model of Literary Language

摘要

We consider the task of predicting how literary a text is, with a goldstandard from human ratings. Aside from a standard bigram baseline, we applyrich syntactic tree fragments, mined from the training set, and a series ofhand-picked features. Our model is the first to distinguish degrees of highlyand less literary novels using a variety of lexical and syntactic features, andexplains 76.0 % of the variation in literary ratings.