[developers] Parsing a profile with ace

Thu Jan 8 04:51:07 CET 2015

The version of ace in Ubuntu+LKB 19, so fairly recent.  It seems I was just
putting the -O in the wrong place...

Emily

On Wed, Jan 7, 2015 at 7:48 PM, Woodley Packard <sweaglesw at sweaglesw.org>
wrote:

> What version of ACE are you using?  This has been around for a while.  Be
> sure you use letter O, not number 0.  Example below.
> -Woodley
>
> $ echo "Dogs bark." | ./ace -g erg-1212.dat -O
> SENT: Dogs bark.
> root 470 0 2 468 root_inffrag
> (468 vp_fin-frg_c 0 2 1 3 [] (455))
> (455 hd-cmp_u_c 0 2 1 3 [] (132,435))
> (132 v_3s-fin_olr 0 1 1 2 [] (44))
> (44 dog_v1 0 1 1 1 [] (32))
> (435 hdn_bnp_c 1 2 1 3 [] (135))
> (135 w_period_plr 1 2 1 2 [] (127))
> (127 n_ms-cnt_ilr 1 2 1 2 [] (34))
> (34 bark_n1 1 2 1 1 [] (29))
> root 471 0 2 454 root_strict
> (454 sb-hd_mc_c 0 2 1 3 [] (422,134))
> (422 hdn_bnp_c 0 1 1 3 [] (130))
> (130 n_pl_olr 0 1 1 2 [] (43))
> (43 dog_n1 0 1 1 1 [] (32))
> (134 w_period_plr 1 2 1 2 [] (126))
> (126 v_n3s-bse_ilr 1 2 1 2 [] (33))
> (33 bark_v1 1 2 1 1 [] (29))
> token 29 1 2 0 token [ +FORM "bark." +FROM "5" +TO "10" +ID *diff-list* [
> LIST *list* LAST *list* ] +TNT null_tnt [ +TAGS *null* +PRBS *null* +MAIN
> tnt_main [ +TAG "NN" +PRB "1.0" ] ] +CLASS alphabetic [ +INITIAL - +CASE
> non_capitalized+lower ] +TRAIT token_trait [ +UW - +IT italics +LB
> bracket_null +RB bracket_null +HD token_head [ +LL ctype [ -CTYPE- string ]
> +TG string ] ] +PRED predsort +CARG "bark" +TICK bool ]
> token 30 0 1 0 token [ +FORM "Dogs" +FROM "0" +TO "4" +ID *diff-list* [
> LIST *list* LAST *list* ] +TNT tnt [ +TAGS *cons* [ FIRST "NNS" REST *null*
> ] +PRBS *cons* [ FIRST "1.0" REST *null* ] +MAIN tnt_main [ +TAG "NNS" +PRB
> "1.0" ] ] +CLASS alphabetic [ +INITIAL + +CASE capitalized+lower ] +TRAIT
> token_trait [ +UW + +IT italics +LB bracket_null +RB bracket_null +HD
> token_head [ +LL ctype [ -CTYPE- string ] +TG string ] ] +PRED
> "_dogs/NNS_u_unknown_rel" +CARG "Dogs" +TICK bool ]
> token 31 1 2 0 token [ +FORM "bark." +FROM "5" +TO "10" +ID *diff-list* [
> LIST *list* LAST *list* ] +TNT tnt [ +TAGS *cons* [ FIRST "NN" REST *null*
> ] +PRBS *cons* [ FIRST "1.0" REST *null* ] +MAIN tnt_main [ +TAG "NN" +PRB
> "1.0" ] ] +CLASS alphabetic [ +INITIAL - +CASE non_capitalized+lower ]
> +TRAIT token_trait [ +UW + +IT italics +LB bracket_null +RB bracket_null
> +HD token_head [ +LL ctype [ -CTYPE- string ] +TG string ] ] +PRED
> "_bark/NN_u_unknown_rel" +CARG "bark" +TICK bool ]
> token 32 0 1 0 token [ +FORM "dogs" +FROM "0" +TO "4" +ID *diff-list* [
> LIST *list* LAST *list* ] +TNT null_tnt [ +TAGS *null* +PRBS *null* +MAIN
> tnt_main [ +TAG "NNS" +PRB "1.0" ] ] +CLASS alphabetic [ +INITIAL + +CASE
> capitalized+lower ] +TRAIT token_trait [ +UW - +IT italics +LB bracket_null
> +RB bracket_null +HD token_head [ +LL ctype [ -CTYPE- string ] +TG string ]
> ] +PRED predsort +CARG "Dogs" +TICK bool ]
> NOTE: 1 readings [forest], RAM: 634k
>
>
>
> NOTE: parsed 1 / 1 sentences, avg 634k, time 0.01739s
>
>
> On Jan 7, 2015, at 6:59 PM, "Emily M. Bender" <ebender at uw.edu> wrote:
>
> Thanks, Woodley.  -O doesn't seem to be an option that ace recognizes, at
> least not as I've tried it...
>
> On Wed, Jan 7, 2015 at 6:55 PM, Woodley Packard <sweaglesw at sweaglesw.org>
> wrote:
>
>> Glad you all have communicated so effectively, even before I am able to
>> get to a real computer!
>>
>> One extra note: to record a full forest profile, pass -f to art and -O to
>> ace.  That's if memory serves, anyway.  More  from a real keyboard when I
>> get home if  you have trouble!
>>
>> -Woodley
>>
>> On Jan 7, 2015, at 6:14 PM, Michael Wayne Goodman <
>> goodmami at u.washington.edu> wrote:
>>
>> > Yes, you can use Art: http://sweaglesw.org/linguistics/libtsdb/art
>> >
>> > $ art -a $ACECOMMAND path/to/profile
>> >
>> > The input profile is not a skeleton, but an empty instantiated
>> > profile. Use the bundled mkprof command to create an empty profile
>> > from a skeleton.
>> >
>> > ACECOMMAND is a string with the command that calls ACE. It might look
>> like this:
>> >
>> > ACECOMMAND="path/to/ace -g path/to/compiled/grammar.dat"
>> >
>> > We've found out recently that you can pipe commands in that string,
>> > which we use for preprocessing:
>> >
>> > ACECOMMAND="preprocessor | ace ..."
>> >
>> > There's more info at the Art website and ACE's
>> > (http://sweaglesw.org/linguistics/ace/). Hope that helps
>> >
>> > On Thu, Jan 8, 2015 at 10:02 AM, Emily M. Bender <ebender at uw.edu>
>> wrote:
>> >> Thanks, Mike.  Looks like those directions involve invoking ace via
>> [incr
>> >> tsdb()] and assume the logon tree.  I can give that shot tomorrow on
>> campus,
>> >> but for now I'm working with Ubuntu+LKB (so without the full logon
>> >> environment).  I thought there was a lighter-weight way of processing
>> >> profiles with ace and storing the results?
>> >>
>> >> Emily
>> >>
>> >> On Wed, Jan 7, 2015 at 5:57 PM, Michael Wayne Goodman
>> >> <goodmami at u.washington.edu> wrote:
>> >>>
>> >>> The URL might be only accessibly within NTU. I attach the text file.
>> >>>
>> >>> On Thu, Jan 8, 2015 at 9:52 AM, Emily M. Bender <ebender at uw.edu>
>> wrote:
>> >>>> Thanks for the quick reply, Sanghoun! Unfortunately, that URL isn't
>> >>>> working
>> >>>> for me ... does it work for you?
>> >>>>
>> >>>> Google Chrome could not load the webpage because 172.21.174.40 took
>> too
>> >>>> long
>> >>>> to respond. The website may be down, or you may be experiencing
>> issues
>> >>>> with
>> >>>> your Internet connection.
>> >>>>
>> >>>> On Wed, Jan 7, 2015 at 5:44 PM, Sanghoun Song <sanghoun at uw.edu>
>> wrote:
>> >>>>>
>> >>>>> Dear Emily,
>> >>>>>
>> >>>>> Last time, Dan gave us the instruction session of using FFTB. Here
>> is
>> >>>>> the
>> >>>>> step-by-step instruction. I hope this will be of help!
>> >>>>>
>> >>>>> http://172.21.174.40/grameng/data/fftb_dan.txt
>> >>>>>
>> >>>>> Sanghoun
>> >>>>>
>> >>>>> On Thu, Jan 8, 2015 at 9:38 AM, Emily M. Bender <ebender at uw.edu>
>> wrote:
>> >>>>>>
>> >>>>>> Dear Woodley (cc: developers),
>> >>>>>>
>> >>>>>> I have a grammar for Matsigenka which is wildly ambiguous (because
>> >>>>>> it's
>> >>>>>> automatically derived...) and I'd like to explore that ambiguity
>> using
>> >>>>>> FFTB.
>> >>>>>> I have an [incr tsdb()] profile, but I can't find any documentation
>> >>>>>> (starting from either AceTop or FftbTop) on how to create a
>> >>>>>> full-forest
>> >>>>>> profile with ace.  Can you point me in the right direction?
>> >>>>>>
>> >>>>>> Thanks,
>> >>>>>> Emily
>> >>>>>>
>> >>>>>> --
>> >>>>>> Emily M. Bender
>> >>>>>> Professor, Department of Linguistics
>> >>>>>> Check out CLMS on facebook! http://www.facebook.com/uwclma
>> >>>>>
>> >>>>>
>> >>>>>
>> >>>>>
>> >>>>> --
>> >>>>> ====================================
>> >>>>> Sanghoun Song
>> >>>>> Ph.D. in Computational Linguistics | http://corpus.mireene.com
>> >>>>> NTU Computational Linguistics Lab. | http://compling.hss.ntu.edu.sg
>> >>>>> ====================================
>> >>>>
>> >>>>
>> >>>>
>> >>>>
>> >>>> --
>> >>>> Emily M. Bender
>> >>>> Professor, Department of Linguistics
>> >>>> Check out CLMS on facebook! http://www.facebook.com/uwclma
>> >>>
>> >>>
>> >>>
>> >>> --
>> >>> -Michael Wayne Goodman
>> >>
>> >>
>> >>
>> >>
>> >> --
>> >> Emily M. Bender
>> >> Professor, Department of Linguistics
>> >> Check out CLMS on facebook! http://www.facebook.com/uwclma
>> >
>> >
>> >
>> > --
>> > -Michael Wayne Goodman
>>
>
>
>
> --
> Emily M. Bender
> Professor, Department of Linguistics
> Check out CLMS on facebook! http://www.facebook.com/uwclma
>
>
>

-- 
Emily M. Bender
Professor, Department of Linguistics
Check out CLMS on facebook! http://www.facebook.com/uwclma
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.delph-in.net/archives/developers/attachments/20150107/e84d95a2/attachment.html>