[CWB] a question regarding the keywords analysis in CQPweb

[Ray Wu]

hi Andrew,

Thanks for your explanation.

What is in my mind is this: suppose there are some neologies in the study corpus (for instance, proper nouns in some political news)  but not present in the reference corpus. According to the current procedure, it means two separate steps have to be involved to find out the keywords.  It is more convenient and consistent to combine these two steps?


Best,
Ray


At 2013-10-31 20:46:41,"Hardie, Andrew" <a.hardie at lancaster.ac.uk> wrote:


I copied this aspect of the user interface. But I imagine I think the rationale was probably: Because if you want words in list 1 that don’t occur at all in list 2, then you’ve got the “compare frequency lists” function on the same screen.

 

Happy to change it if there’s a general feeling that being able to go down to zero would be useful.

 

best

 

Andrew.

 

From:cwb-bounces at sslmit.unibo.it [mailto:cwb-bounces at sslmit.unibo.it] On Behalf Of Ray Wu
Sent: 31 October 2013 12:29
To:cwb at sslmit.unibo.it
Subject: [CWB] a question regarding the keywords analysis in CQPweb

 

Dear members,

It is known that currently CQPweb sets a "Min freq"  constraint for keyword analysis.  Isn't it possible that "Min freq (list 2):" could be zero in some cases? Why should the lowest number of "Min freq (list 2):" be set to 1 instead of zero?

Thanks for any pointers to relevant literature.

Best,
Ray

 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://devel.sslmit.unibo.it/pipermail/cwb/attachments/20131031/5d86c5bc/attachment.html>