Laboratory of Computer and Information Science / Neural Networks Research Centre CIS Lab Helsinki University of Technology

Competition 1: Turkish


Table: The submitted unsupervised morpheme analysis compared to the Gold Standard in Turkish (Competition 1).
AUTHOR METHOD PRECISION RECALL F-MEASURE
Monson Paramor+Morfessor 51.88% 52.10% 51.99%
Monson Paramor 56.67% 39.42% 46.50%
Monson Morfessor 73.92% 26.06% 38.53%
Morfessor MAP 76.36% 24.50% 37.10%
Zeman 1 65.81% 18.79% 29.23%
Morfessor baseline 89.20% 11.32% 20.08%
Kohonen et al - 93.25% 6.15% 11.53%
Zeman 3 73.30% 3.01% 5.79%


Table: Corresponding results in the Morpho Challenge 2007 (for reference)
AUTHOR METHOD PRECISION RECALL F-MEASURE
Morfessor MAP 76.36% 24.50% 37.10%
Zeman - 65.81% 18.79% 29.23%
Bordag 5a 81.31% 17.58% 28.91%
Bordag 5 81.44% 17.45% 28.75%
Bernhard 2 73.69% 14.80% 24.65%
Bernhard 1 78.22% 10.93% 19.18%
McNamee 3 65.00% 10.83% 18.57%
McNamee 4 85.49% 6.59% 12.24%
McNamee 5 94.80% 3.31% 6.39%

Return to the result page

HOME | RULES | SCHEDULE | DATASETS | EVALUATION | WORKSHOP | RESULTS | FAQ | CONTACT

You are at: CIS → Unsupervised Morpheme Analysis -- Morpho Challenge 2008

Page maintained by webmaster at cis.hut.fi, last updated Wednesday, 09-Jul-2008 17:25:10 EEST