Hi Karl,
Could you elaborate on the kind of features you would use to train
Markov chains? I worked some on CRF's (Conditional Random Fields) for one of
the information extraction projects. It would be useful to know your approach
as well as I might be able to pitch in some ideas.
Also, I couldn't find the references listed in the email by [EMAIL
PROTECTED] below. Could you please repost them?
Thanks,
Rajesh Munavalli
http://munavalli.blogspot.com
-----Original Message-----
From: karl wettin [mailto:[EMAIL PROTECTED]
Sent: Tue 4/11/2006 8:19 AM
To: java-dev@lucene.apache.org
Subject: Re: Contextual suggestions (for spelling)
11 apr 2006 kl. 02.20 skrev [EMAIL PROTECTED]:
>
>> Any comments on this? Questions?
>
> You've sketched a slightly degenerate version of Shannon's
> noisy channel model. Here are three refs to help
> you sort out the usual approach:
Wow! Thanks!
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
