FREQ

[Brian MacWhinney]

Dear ChiBolts,
    I am replying here to a message that Darinka posted to info-childes.  Because this is a message about the functioning of CLAN programs, it is best discussed on chibolts instead.  Here is her message and my answer:

> Dear collegues,
>  
> I wanted to match codes on the %mor tier to words on the main line and I used MODREP for it:
>  
> MODREP +u +k +r2 +b%mor +c* +t at ID=2.*.*.*.*.*  +o"v|*" *.mor
>  
> But then I found out that the output of this command do not provide the same frequency for codes as I get from FREQ on %mor tier. MODREP provided systematically larger frequency then FREQ.
>  
> I had expected them to be exactly the same. Do I missunderstand something?
>  
>  
> Darinka Andjelkovic
>  
> Laboratory of Experimental Psychology and Institute of Psychology
> Faculty of Philosophy
> University of Belgrade
> Serbia

Dear Darinka,
     The difference arises from the fact that the %mor line excludes repetitions and several other things from the main line.  The goal of the %mor line is to analyze morphology and syntax.  To do this, excluding repetitions helps in automatic part of speech tagging.  The %mod line, on the other hand, is pegged to the main line in a way that includes all words, including repetitions.  Therefore, it will have larger numbers than a FREQ on the %mor line.  You can modify what FREQ looks at on the main line by using the +r6 switch to exclude repetitions.  You might check that.  In general, few people have used the %mod line as the target for FREQ, so we haven't payed much attention to the details of what is there.  

-- Brian MacWhinney

-- 
You received this message because you are subscribed to the Google Groups "chibolts" group.
To post to this group, send email to chibolts at googlegroups.com.
To unsubscribe from this group, send email to chibolts+unsubscribe at googlegroups.com.
For more options, visit this group at http://groups.google.com/group/chibolts?hl=en.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/chibolts/attachments/20120207/89d3324f/attachment.htm>