abydos icon indicating copy to clipboard operation
abydos copied to clipboard

Published 20 hours ago • verified publisher icon chrislit

unimplemented tasks for review

Open chrislit opened this issue 5 years ago • 0 comments

These are things that were considered for implementation, but were not. Some don't make sense. Some looked too hard/impossible to implement or implement efficiently.

They may be worth further consideration.

Distance measures:

  • [ ] Model Distance, CMD, CMC (10.1016/j.aca.2009.06.032) couldn't figure out how to implement this in a way that makes sense -- I'm getting large negative r_AB and hence distances
  • [ ] Normalized Levenshtein from 10.1109/34.232078 unneeded, only applies to edit distances with weight matrices
  • [ ] https://people.eng.unimelb.edu.au/jzobel/fulltext/cikm05mbcmz.pdf omit
  • [ ] METEOR https://en.wikipedia.org/wiki/METEOR punted
  • [ ] NIST score https://en.wikipedia.org/wiki/NIST_(metric) punted
  • [ ] Sorted Winkler (cf. https://digitalcollections.anu.edu.au/bitstream/1885/44521/3/TR-CS-06-02.pdf)

chrislit avatar Feb 15 '19 22:02 chrislit