Comment on Simple Haskell webapp: Generate random tweets using Markov chainsparentComments−dododo15yare you using smoothing for large n? kneser-ney smoothing seems to give the best results.http://nlp.stanford.edu/~wcmac/papers/20050421-smoothing-tut...−ma2rten15yI did not read the paper, but what does more accurate mean in this case ? Likelihood of some unseen data ? Seems pretty hard to define or measure to me, if the goal is shear amusement.−jaspervdjOP15yInteresting link, thanks!
Comments
are you using smoothing for large n? kneser-ney smoothing seems to give the best results.
http://nlp.stanford.edu/~wcmac/papers/20050421-smoothing-tut...
I did not read the paper, but what does more accurate mean in this case ? Likelihood of some unseen data ? Seems pretty hard to define or measure to me, if the goal is shear amusement.
Interesting link, thanks!