[Morfessor-announce] Morfessor FlatCat released

[Mailing list for Morfessor Announcements]

Hi,

We are excited to reveal Morfessor FlatCat, a new morphological
segmentation method in the Morfessor family.

Morfessor FlatCat has now been officially published at
http://www.cis.hut.fi/projects/morpho/
The source code is available at
https://github.com/aalto-speech/flatcat

Morfessor FlatCat combines the semi-supervised training previously used in
Morfessor Baseline, with the morph categories from Morfessor Categories-MAP
and Categories-ML. There are four morph categories: three proper categories
(stem, prefix and suffix) and a non-morpheme category (marked ZZZ), for
segments that are not proper morphs because they are fragments of a larger
morph. The morph categories allow Morfessor FlatCat to be used as a
stemmer.

The name Morfessor FlatCat comes from the use of a lexicon without 
hierarchical structure (a flat lexicon), together with the morph
categories.

For a detailed description of the method, see the paper:
Grönroos, S.-A., Virpioja, S., Smit, P., and Kurimo, M. (2014).
Morfessor FlatCat: An HMM-based method for unsupervised and semi-supervised
learning of morphology.  In proceedings of the 25th International
Conference on Computational Linguistics.  Pages 1177-1185, Dublin, Ireland,
August 2014, Association for Computational Linguistics.
http://www.aclweb.org/anthology/C/C14/C14-1111.pdf

Morfessor FlatCat will be presented at the COLING conference in Dublin,
Ireland next week. Find us at Poster session II (Tuesday 26th August
14:00-15:15) in the 2nd Floor Lobby.

http://www.coling-2014.org/schedule/tuesday26august.php

--

Stig-Arne Grönroos
Doctoral student
Aalto University