Hi,
I've developped a collocation extractor based on statistics
that takes as input file any text in any language and produces
a list of (contiguous and non-contiguous) collocation candidates.
It is based on positional ngrams, Mutual Expectation (association
measure) and the GenLocalMaxs algorithm (acquisition process).
It is acessible at http://senta.di.ubi.pt. You may register and
send your text to SENTA to automatically receive the results by email.
The processing is done in loco so that small text uploads are
compulsory.
For any question or comment, feel free to contact me.
Gaël.
-- ------------------------------------------------------- Gaël Harry Dias - Assistant Professor Beira Interior University - Computer Science Department 6200-053 - Covilhã - Portugal http://www.di.ubi.pt/~ddg - ddg@di.ubi.pt -------------------------------------------------------
This archive was generated by hypermail 2b29 : Wed May 14 2003 - 14:31:33 MET DST