Philgoo Han writeup of Cohen, Kautz and McAllester

This is a review of Cohen_2000_hardening_soft_information_sources by user:Ironfoot.

• This paper covers entity recognition over segmented text(record)
  • Finding same entities in different appearing
  • Token distance measure: I_pot - any experiments on which measure is optimal?
  • Maximizing joint probability
    • Greedy approximation algorithm
• With appropriate distance measure this will be extendable to various hardening information
  • Maybe this is why a single distance measure was not proposed
• It would have been good to show test results on various data with various distance measure