How to find relative clauses in CLEC?

[asan82]

I want to study the use of relative clasues by Chinese learerns of English.But the question is how to find these examples in CLEC. Can anyone help me?

[csli]

1. Tag a CLEC file with CLAWS;
2. Use *_CST (and other tags) in a concordancer (e.g. WordSmith) to find the relative clauses.

[tiger]

but tagging the whole corpus is too time-consuming and tedious for a single person. is there a better way to do this?

[xiaoz]

No other reliable way, unless you want to read through the whole corpus to identify relativisers.

[xujiajin]

Can brill tagger/Gotagger tag the relative clauses, since many of us do not have CLAWS?

[清风出袖]

http://www.comp.lancs.ac.uk/computing/research/ucrel/claws/trial.html offers trial service of CLAWS.

[清风出袖]

He_PPHS1 is_VBZ the_AT man_NN1 that_CST I_PPIS1 love_VV0 ._.
is the result of CLAW C7. yet i don't know which is the tagging of the relative clause.


[本贴已被 作者 于 2005年09月25日 23时46分10秒 编辑过]

[xiaoz]

Have anoyone tried the Monty Tagger:
http://web.media.mit.edu/~hugo/montylingua/index.html

[清风出袖]

it looks like it is created by a Chinese programmer. I will have a try! Are there any advantages over CLAW in the tagging that the Monty Tagger has?

[动态语法]

Unless you have a TreeBank style corpus (e.g. LDC, ICE-GB, etc.), you will
have to put in some manual labor into it. When the object of inquiry is complex
there is no way that you can do it easy.

Incidentally, Prof Cathy Ball of Georgetown University has an excellent tutorial for
finding relative clause markers for a gender-based study:

http://www.georgetown.edu/faculty/ballc/corpora/tutorial3.html#RTFToC31
(Note: please copy the whole line above to get to #RTFToC31 directly.)

Note that there are zero-marked RCs, for which there is no way you can do without
looking at them manually.

Some of her techniques may be improved these days with more effecient automated
or semi-automated procedures, however.

以下是引用 tiger 在 2005-9-25 22:15:52 的发言：
but tagging the whole corpus is too time-consuming and tedious for a single person. is there a better way to do this?

[asan82]

Thank you all for the kind suggestions.
I've read Prof Cathy Ball's case study and was amazed at all the manual work it had taken.
Perhaps it's better for me to narrow my research scope or shift the focus.

[xiaoz]

If you have a CLAWS tagged English corpus and WordSmith, occurrences of that deletions can be extracted reasonably reliably without much manual work. You can download the search algorithms and the zipped archive for file-based search at
http://bowland-files.lancs.ac.uk/corplang/cbls/resources.asp

[asan82]

怎么打不开？

[xiaoz]

打不开what?

以下是引用 asan82 在 2005-9-27 18:40:35 的发言：
怎么打不开？

[asan82]

链接。
可能是网络问题，我再试试看。

[shanhu]

请问我要是想写衔接手段失误该怎么检索呢

[xujiajin]

File-based concordancing will help.

http://www.corpus4u.com/forum_view.asp?view_id=228&forum_id=7

[tiger]

以下是引用 asan82 在 2005-9-26 21:43:30 的发言：
Thank you all for the kind suggestions.
I've read Prof Cathy Ball's case study and was amazed at all the manual work it had taken.
Perhaps it's better for me to narrow my research scope or shift the focus.


what about the use of relativisers such as that and which?