[Haskell-cafe] The best way to call Java from Haskell?
creswick at gmail.com
Tue Oct 11 18:44:13 CEST 2011
On Tue, Oct 11, 2011 at 1:56 AM, dokondr <dokondr at gmail.com> wrote:
> On Tue, Oct 11, 2011 at 12:42 PM, Yitzchak Gale <gale at sefer.org> wrote:
>> Just out of curiosity, why do you not consider GF
>> at all similar? To an outsider like me, there does
>> appear to be quite a bit of similarity.
> As I understand GF is well suited for parsing well defined formal languages.
> Not sure that GF can be used as NLP parser for blog messages that I need.
That is correct - more or less. GF is a very expressive language, and
it can handle a great deal of natural language, but it /does/ require
that the input be grammatically correct, and it is difficult to work
in unexpected vocabulary.
GF is fantastic for making flexible Controlled Natural Languages, and
it excels at producing human-readable text, but it is an entirely
different beast from a statistical natural language parser, such as
re: the original question --
The best method I've found for interfacing Haskell / Java for NLP is
to share data with some common format, such as UIMAs CAS
serialization. We really ought to start up a group of people
interested in that over on the Haskell NLP list and see what we have
if we pool all our efforts.
> Please correct me if I am wrong.
> As a general note, Java has tons of useful libraries that will take infinite
> time to re-implement in Haskell. To my mind it makes a lot of sense to have
> a reliable mechanism to call Java from Haskell.
> BTW, yet another way to do this: wrap Java library in RESTFUL web service )
> All the best,
> Dmitri O. Kondratiev
> "This is what keeps me going: discovery"
> dokondr at gmail.com
> Haskell-Cafe mailing list
> Haskell-Cafe at haskell.org
More information about the Haskell-Cafe