Key n-grams

[alex boulton]

Dear Laurence & all

I know a work-around which Laurence showed me for finding key n-grams in AntConc v3, but I don't seem to be able to use the same procedure with v4. Does anyone know how to do this?

Best wishes

alex

[Laurence Anthony]

Hi Alex,

Here are the steps I followed assuming you are comparing Corpus A with Corpus B. I used the AmE06 learned sub-corpus as Corpus A and the full corpus as Corpus B.

1) Open Corpus A as a target corpus.

2) In the N-gram tool, generate an n-gram list (e.g. size 2) making sure that the page size is "All hits"

3) In the File menu, save the results using the "Save Current Tab Results"

4) Repeat steps 1-3 for Corpus B.

5) In the corpus manager, use the "word list" option to load the Corpus A n-gram list as a word list corpus. Then repeat for the Corpus B n-gram list.

6) In the corpus manager, load the Corpus A n-gram list corpus as the target and the Corpus B n-gram list corpus as the reference.

7) In the Keyword tool, hit start and you should get results looking like the following:

Using the same approach, it should be possible to create key collocates, key clusters and so on. 

Saying that, I think this a pretty clumsy set of steps that are needed to do something relatively simple. I'm currently working on AntConc 4.4 (currently in beta release). I think this is a good time to add a much more elegant way to generate keywords for the different tools (clusters, n-grams, collocates). My idea has long been to have an option in the keyword preferences (or maybe in the main interface) to choose the *type* of keywords you want, with "words" being the default, but having a dropdown for the other tools. The tricky part is showing the user what is happening. If they choose "Key N-grams" as the type of keywords, the user also needs to specify the size of the n-gram, and with key collocates, the user needs to specify the collocate windows size. The easy option is to simply have the system 'mirror' the current settings that the user has for the n-grams or collocates, and generate the same results on the reference corpus. But, keeping that transparent is not so straightforward. If you (or anybody else) have any ideas for this, do let me know!

Regards,

Laurence.

###############################################################
Laurence ANTHONY, Ph.D.
Professor of Applied Linguistics
Faculty of Science and Engineering
Waseda University
3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan
E-mail: antho...@gmail.com
WWW: http://www.laurenceanthony.net/
###############################################################

--
You received this message because you are subscribed to the Google Groups "AntConc-Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to antconc+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/antconc/dd67a109-73cd-4220-991a-1a7faa43bea4n%40googlegroups.com.