Bayesian Classifiers for WSD

• An example involving the verb 'cut' (here's another example involving document categorization)
  Let's assume we have three senses to classify:
  • cut(1): to produce a separation in the material of an object (e.g., 'cut bread', 'cut one's finger'
  • cut(2): to remove something from an (information-bearing) object (e.g., 'cut a paragraph', 'cut a scene')
  • cut(3): to remove a player from a team
• How common is each of these senses compared to the others?
• What features of the context around an occurrence of 'cut' would help us choose the correct sense?

(2)

(continue with example)

(return to syllabus)