Supervised Categorization for Habitual versus Episodic Sentences

  title={Supervised Categorization for Habitual versus Episodic Sentences},
  author={Thomas Mathew and E. Graham Katz},
We implement and evaluate systems for automatically distinguishing habitual and episodic sentences. Using features such as tense, aspect, noun phrase characteristics, temporal modifiers, specific adverb modifiers, and specific verb auxiliaries on genericity we built and evaluated a supervised machine learning classifier that provides 86.3% precision in disambiguating habitual and episodic sentences. This compares against a majority class baseline of 73.1% precision. In order to support these… CONTINUE READING