Extract word list from a mor tagged corpus, restore the original form of the tagged words
ihappylearning at gmail.com
Thu Dec 20 12:29:52 UTC 2012
Dear Mr. MacWhinney,
Thank you so much for your quick response and the information. What I was
trying to do is to make lists of words based on the Part of Speech of the
texts, to do collocation study. For example, to extract all the adjectives
and adverbs from two text corpuses (say C1, and C2), and then matching the
adj+adv collocation phrases in other text corpus (C3), and to measure the
occurring frequency; or vice versa, to extract adj+adv collocation phrases
in C3 and see how frequent these adj and adv words occur in C1 and C2.
I use CLAN to mor the text corpus, and then I use a regular expression
software to match the words in the mor lines. For example,
I use the following two regular expressions to match and extract all the
adjective words in a mor taggaed corpus:
the result of extraction looks like this:
And then, I use regular expression again to remove the head 'adj\|' and
'adj\:\w+' at once:
And then I restore them manually as the following:
The problem is that when the list becomes large, it is almost impossible to
check them by hand. That's why I was wondering if there is any better way
of doing this, i.e. to restore the word form, or to extract words or
phrases by the part of speech of the text?
Brian MacWhinney於 2012年12月20日星期四UTC+8上午6時15分04秒寫道：
> Dear YuJen,
> MOR is designed to change the forms on the main line to analyzed forms
> on the %mor line. CLAN is not really designed to give you access to both
> the main line form and the %mor line form at the same time. It is not
> clear from your message exactly when and how you need to have access to the
> main line forms. There are some limited abilities to retrieve
> corresponding forms that can work in COMBO or KWAL, but the best way to
> approach this systematically would be through a search of the XML version
> of the corpus where the %mor forms and the main line forms are together in
> a single bundle. You can create XML versions of CHAT corpora using the
> Chatter program. Then you would use your regular expression matching
> software to process the XML.
> --Brian MacWhinney
> Dear all,
> I have question about how to restore the word form from a mor tagged
> I was making word lists according to part of speech. I used mor to tag a
> text corpus, and then I use a regular expression software to extract the
> words by matching the part of speech tags created by mor. The word list
> extracted seems fine, but I found that some of the lexical forms have
> changed by mor in the %mor tier. For example, the plural form of a noun,
> the tenses of a verb...
> I was wondering is there anyway that I can restore them to the original
> form in CLAN? Or other efficient methods without needing to check and
> modify them one by one.
> Thank you.
> Huang, YuJen
> You received this message because you are subscribed to the Google Groups
> "chibolts" group.
> To unsubscribe from this group, send email to
> To view this discussion on the web visit
> For more options, visit https://groups.google.com/groups/opt_out.
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.
To view this discussion on the web visit https://groups.google.com/d/msg/chibolts/-/zSAqejP0OkgJ.
For more options, visit https://groups.google.com/groups/opt_out.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Chibolts