[pet] State of the project?

[Francis Bond]

G'day,

I am a user of PET, mainly using it with the English and Japanese
grammars (ERG and Jacy).

> I am totally new to PET.  Just this week I got it to build on mac os x 10.6
> and was able to run some tests.  I have been working with other parsers --
> the Stanford PCFG-derived tool, the Malt dependency parser, the link grammar
> device -- and found that PET seems more robust and more prone to returning
> good trees compared to the others.  Frankly I was surprised because, at
> least in my circles (US East Coast computational linguistics) I don't think
> PET is yet a household name.

How nice to hear.  Until recently, PET has had some issues with
coverage (on open news text it is typical for 10%+ of sentences not to
get a parse), which I think has  put many potential users off.  There
has been considerable improvement recently, and more is planned.

> I'm therefore curious about the current state of the project.  I read the
> history on a wiki, and hopped around the mail lists, but wasn't able to get
> a sense of whether there is a significant community of developers and users.
>  Is there a fair degree of forward momentum and enthusiasm?

Yes and yes.  Because of the considerable interaction with other
systems, a lot of discussion takes place on the developers list (and
off-line at the DELPH-IN Summit).

> Given the quality of the system, I hope so.
> Also, if one wanted to volunteer a bit as a programmer (work + dissertation
> mean not a lot of time available), where might one start?  Tickets on the
> Trac?

Welcome aboard.  I am not the best person to answer this question, but
Trac seems like a reasonable start.


-- 
Francis Bond <http://www3.ntu.edu.sg/home/fcbond/>
Division of Linguistics and Multilingual Studies
Nanyang Technological University