Difference between revisions of "Nschneid writeup of Borkar 2001"

From Cohen Courses
Jump to navigationJump to search
 
m (1 revision)
 
(No difference)

Latest revision as of 10:42, 3 September 2010

This is Nschneid's review of Borkar_2001_Automatic_Segmentation_of_Text_Into_Structured_Records

I think it's interesting how they nest the HMMs in this paper. Is this a standard technique? Is it equivalent to another more convenient formulation of the problem (e.g. append the two types of states together)?

  • I'm not sure how widely known this is, but there are standard ways of composing transducers, I believe this sort of nesting is one of the operations in the AT&T package by Pereira & Riley,eg. - Wcohen 14:21, 24 September 2009 (UTC)