gestures coded without objects, and obtaining parser input

Mon Oct 23 04:55:08 UTC 2017

Thanks for the quick response!   I don't think that porting to Unix should 
be necessary.  I could use another OS.   I also now see that FLO might be 
sufficient.   Some example output from FLO might illustrate what I was 
trying to say. 

Here's an example sentence:

*PAR:    and they whole [: all] [* s:r] <mushed up into> [//]
    &=hands:together &=laughs mushed in together .  
%mor:    coord|and pro:sub|they post|all v|mush-PAST adv|in adv|together .

I think when I said "what the parser sees", this was motivated by the fact 
that the %mor line reflects the dependency parse after a lot of 
postprocessing (e.g., removal of repetition, replacement with annotator's 
suggested word, etc).  If I were to use an external parsing program, I 
would want it to see the results of that postprocessing, rather than the 
original text with all its annotations.  And it looks like FLO might be 
getting something close to that.

Thanks again for your help -- 
-Karin

On Sunday, October 22, 2017 at 11:24:51 PM UTC-4, Leonid Spektor wrote:
>
> Karin,
>
> The "chat2..." commands convert CHAT formatted data to format of the 
> corresponding applications "Anvil", "Elan", "Praat", "EXMARaLDA" and "CA". 
> CHAT2CONLL was not ported to Unix. It can be ported if people in charge at 
> our end approve it, but I would suggest that you try that command in either 
> MacOS or Windows OS version of CLAN to see if it creates the output that 
> you really want before we spent time porting it to Unix. As far as I know, 
> no one except us in house have ever used CHAT2CONLL command.
>
> I have to ask you to clarify what do you mean by "text that is identical 
> or similar to what CLAN's parser sees". CLAN has one parser MOR and it sees 
> the CHAT formatted files, so they do not need to be converted to anything 
> else. Perhaps I misunderstood your question and so I gave you the wrong 
> suggestion to use CHAT2CONLL. Please, explain in more details what it is 
> you are trying to achieve and what data format is your starting point. It 
> would also help if you could give an example of what output format you want 
> to get.
>
>
> Leonid.
>
> On Oct 22, 2017, at 17:53, Karin <cox.k... at gmail.com <javascript:>> wrote:
>
> Quick follow-up question on this discussion from a month ago: I do not see 
> chat2conll in the list of available programs in unix-clan/unix/bin.  Here's 
> a list of all files that start with "chat2": 
>
> chat2anvil  chat2ca  chat2elan  chat2praat  chat2xmar
>
> Would any of those programs create text that is identical or similar to 
> what CLAN's parser sees?   If not, is there a way I can install the 
> chat2conll program (or am I looking in the wrong directory)?
>
> Thanks again -- 
>
> - Karin
>
>
>
>
> On Saturday, September 23, 2017 at 6:37:08 PM UTC-4, Leonid Spektor wrote:
>>
>> Karin,
>>
>> I can only answer your second question. There are two CLAN commands that 
>> will cleanup speaker tiers in CHAT files. It depends on how much cleanup 
>> you need. First command is FLO. It removes all codes and more leaving as 
>> little as just a plain text lines. Second command is CHAT2CONLL. It creates 
>> output in format suited for either one of the following parsers: Depparse, 
>> MaltParser, TurboParser, AnCoraCorpus, Connexor, Clearparser.
>>
>>
>> Leonid.
>>
>> On Sep 23, 2017, at 16:12, Karin <cox.k... at gmail.com> wrote:
>>
>> Hello all,
>>
>> I will list here two questions, but if it's better to split them up 
>> between posts, please let me know.
>>
>> (1) I am wondering how to interpret the following coding of a gesture: 
>>
>> you take <two pieces> [//] &=ges two [/] two &=fingers:two &sli slices 
>> of bread . 
>>
>>
>> Does this imply that (a) the speaker made an empty gesture, and then said 
>> "two", or (b) the annotator forgot a colon?
>>
>>
>> (2) Is there a program in CLAN that allows for the cleaning of the 
>> speaker tiers, such that they appear in the way that the parser sees 
>> them?   If not, is there a CLAN program that I could modify to produce such 
>> output?  I am reasonably experienced with programming, and hopefully would 
>> know enough C++ to manage this. 
>>
>>
>> Thank you in advance for your time --
>>
>> -- Karin
>>
>>
>>
>>
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "chibolts" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to chibolts+u... at googlegroups.com.
>> To post to this group, send email to chib... at googlegroups.com.
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/chibolts/1c48fc40-b165-4d03-980f-46b79ae102d9%40googlegroups.com 
>> <https://groups.google.com/d/msgid/chibolts/1c48fc40-b165-4d03-980f-46b79ae102d9%40googlegroups.com?utm_medium=email&utm_source=footer>
>> .
>> For more options, visit https://groups.google.com/d/optout.
>>
>>
>>
> -- 
> You received this message because you are subscribed to the Google Groups 
> "chibolts" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to chibolts+u... at googlegroups.com <javascript:>.
> To post to this group, send email to chib... at googlegroups.com 
> <javascript:>.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/chibolts/4b8059bc-a3cf-492a-8f9b-e885b827d325%40googlegroups.com 
> <https://groups.google.com/d/msgid/chibolts/4b8059bc-a3cf-492a-8f9b-e885b827d325%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
> For more options, visit https://groups.google.com/d/optout.
>
>
>

-- 
You received this message because you are subscribed to the Google Groups "chibolts" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chibolts+unsubscribe at googlegroups.com.
To post to this group, send email to chibolts at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/f4763e06-d2dd-4c24-a8f0-2612392cc761%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/chibolts/attachments/20171022/015047ca/attachment.htm>