We are happy to announce that the Python implementation of Morfessor,
Morfessor 2.0, has now been officially published at
Morfessor, originally developed by Mathias Creutz and Krista Lagus, is
a method for unsupervised morphological segmentation. Its first
implementation, Morfessor 1.0, was published already in 2005. Since
then, Morfessor has been used to improve processing of morpholigically
complex languages in a wide range of applications such as speech
recognition, information retrieval and machine translation.
Morfessor 2.0 includes a number of new features, including
semi-supervised training and several speed enhancements. The
accompanying Technical Report can be found at
As more Morfessor news are coming up, we invite you to join new
announcement mailing list, firstname.lastname@example.org.
For any questions and comments, contact morfesso[эт]cis.hut.fi.