W02-0704 0.4 points lower accuracy than supervised segmentation . The proposed algorithm can
P14-1010 large advantage to using MADA 's supervised segmentation in this scenario . 5.1 Ablation
P14-1128 learning constraints that guide a supervised segmentation model toward a better solution
P06-2056 such an important language that supervised segmentation methods are already very mature
P09-1012 per character than that based on supervised segmentations . We believe this will be particularly
D11-1056 non-parametric models can complement supervised segmentation . 3 Japanese Noun Phrase Segmentation
W12-6315 constitutes the major drawback of supervised segmentation . In contrast , unsupervised
P09-1012 classics , or unknown languages where supervised segmentation data is difficult or even impossible
D15-1142 also used such mappings to bias a supervised segmentation model toward a better solution
W10-4131 unsupervised . Particularly , supervised segmentation methods can achieve a very high
P12-1025 integrate heterogeneous models . Supervised segmentation and tagging can be improved by
D11-1056 morphological an - alyzer . Although supervised segmentation is very competitive , we showed
W06-3103 Habash and Rambow , 2005 ) three supervised segmentation methods are introduced . However
D13-1061 explored the feasibility of enhancing supervised segmentation by informing the supervised learner
D14-1095 10-fold crossvalidation for the supervised segmentations , while we use the entire set
W13-4402 It was initially intended for supervised segmentation so each corpus is divided between
P09-1012 The result also shows that the supervised segmentations are suboptimal with respect to
D14-1095 algorithms , ranging from near-perfect supervised segmentations to random segmentations , along
W14-2808 achieve any error reduction using a supervised segmentation approach , even though it is
W02-0704 segmentation is equally likely . 2 . Supervised Segmentation ( SUP ) This case also stops
hide detail