I am looking for such a corpus.
In this corpus, text segments that state REASONs, and those text segments that state corresponding EVENTs are tagged.
(Or, just REASONs are tagged)
I want to analyse some causal pattern (such as "because", "therefore", "so", etc.)
in order to get their frequency of occurrence and their precision.
I know Penn Treebank has a tag PRP for purpose and reason.
Besides that, did anyone know other similar corpus?
THANK YOU ALL
This archive was generated by hypermail 2b29 : Wed Mar 19 2003 - 21:32:30 MET