Compost English

Compost English is a tool which combines the Morphium morphological analyzer and the Morce tagger using an innovative semi-supervised training method. The resulting tagger gives the best accuracy achieved for English (on standard PTB data set) so far: 97.43 %

Compost English is written in Perl and C and is available for registered users only. However, the default package does not contain the Morce tagger source code, which you can dowload directly from the Morce website.

Authors: Drahomíra "johanka" Spoustová, Jan Hajič, Jan Raab, Miroslav Spousta..
This work was funded in part by the Companions project sponsored by the European Commission as part of the Information Society Technologies (IST) programme under EC grant number IST-FP6-034434, MSM0021620838, ME838 and LC536 of Ministry of Education, Youth and Sports of the Czech Republic, GA405/06/0589 and GD201/05/H014 of the Grant Agency of the Czech Republic, 1ET101120503 of the Information Society Programme of the National Research Programme of the Czech Republic.
2008 - 2011 © Institute of Formal and Applied Linguistics. All Rights Reserved.