0.25
0.5
0.75
1.25
1.5
1.75
2
Token and Type Constraints for Cross-Lingual Part-of-Speech Tagging
Published on Oct 02, 20132465 Views
We consider the construction of part-of-speech taggers for resource-poor languages. Recently, manually constructed tag dictionaries from Wiktionary and dictionaries projected via bitext have been use
Related categories
Chapter list
Token and Type Constraints for Cross-Lingual Part-of-Speech Tagging00:00
Latent variable CRF model00:07
Part-of-Speech Tagging01:33
Coarse “Universal” Tags - 102:04
Coarse “Universal” Tags - 202:24
Coarse “Universal” Tags - 302:29
Coarse “Universal” Tags - 402:35
Constrains02:53
Models03:11
Complete Token Supervision03:57
Projected Token Constraints - 104:29
Projected Token Constraints - 205:47
Projected Token Constraints - 306:06
Type Constraints 06:37
Type Constraints (HMM) - 107:50
Type Constraints (HMM) - 208:00
Token and Type Constraints - 108:12
Token and Type Constraints - 209:00
Token and Type Constraints - 309:13
Token and Type Constraints - 409:24
Token and Type Constraints - 509:32
Token and Type Constraints - 609:41
Token and Type Results 09:57
Comparing with the State of the Art12:26
Projection Errors - 113:25
Projection Errors - 213:46
Projection Errors - 313:56
Projection Errors - 413:59
Projection Errors - 514:00
Projection Errors - 614:03
Pruning Errors - 114:25
Pruning Errors - 214:39
Pruning Errors - 314:57
Pruning Errors - 416:39
Conclusion16:48
Thanks18:11