Morfessor-enriched features and multilingual training for canonical morphological segmentation

In our submission to the SIGMORPHON 2022 Shared Task on Morpheme Segmentation, we study whether an unsupervised morphological segmentation method, Morfessor, can help in a supervised setting. Previous research has shown the effectiveness of the approach in semisupervised settings with small amounts of labeled data. The current tasks vary in data size: the amount of word-level annotated training data is much larger, but the amount of sentencelevel annotated training data remains small. Our approach is to pre-segment the input data for a neural sequence-to-sequence model with the unsupervised method. As the unsupervised method can be trained with raw text data, we use Wikipedia to increase the amount of training data. In addition, we train multilingual models for the sentence-level task. The results for the Morfessor-enriched features are mixed, showing benefit for all three sentencelevel tasks but only some of the word-level tasks. The multilingual training yields considerable improvements over the monolingual sentence-level models, but it negates the effect of the enriched features.

PDF Abstract

Datasets


Results from the Paper


 Ranked #1 on Morpheme Segmentaiton on UniMorph 4.0 (f1 macro avg (subtask 2) metric)

     Get a GitHub badge
Task Dataset Model Metric Name Metric Value Global Rank Benchmark
Morpheme Segmentaiton UniMorph 4.0 AUUH_E f1 macro avg (subtask 2) 73.21 # 7
lev dist (subtask 2) 31.05 # 10
Morpheme Segmentaiton UniMorph 4.0 AUUH_D f1 macro avg (subtask 2) 72.75 # 8
lev dist (subtask 2) 36.38 # 13
Morpheme Segmentaiton UniMorph 4.0 AUUH_C f1 macro avg (subtask 2) 70.76 # 9
lev dist (subtask 2) 35.94 # 11
Morpheme Segmentaiton UniMorph 4.0 AUUH_B f1 macro avg (subtask 2) 89.77 # 1
lev dist (subtask 2) 3.50 # 1
Morpheme Segmentaiton UniMorph 4.0 AUUH_A f1 macro avg (subtask 2) 89.00 # 2
lev dist (subtask 2) 4.08 # 2
Morpheme Segmentaiton UniMorph 4.0 Bidirectional GRU + Morfessor features (AUUH_F) macro avg (subtask 1) 93.72 # 6
f1 macro avg (subtask 2) 66.73 # 10
lev dist (subtask 2) 36.35 # 12

Methods


No methods listed for this paper. Add relevant methods here