D09-1102 big as possible . In this way , label propagation tends to happen within the same
D08-1061 assigning labels to nodes as graph label propagation . We are given a score ( C ,
D09-1102 anaphoricity of noun phrases via a label propagation algorithm to improve learning
D10-1017 distributions . After running label propagation ( graph propagate ) , the posteriors
D08-1061 Adsorption algorithm . The Adsorption label propagation algorithm ( Baluja et al. , 2008
D09-1102 achieved . Figure 1 presents the label propagation algorithm . Assume : Y : the
D10-1017 semisupervised training without label propagation . This is different from plain
D10-1017 learning and also standard graph label propagation machinery . Graph regularizer
D10-1017 obtained from the graph after label propagation will have lost most of the sequence
D11-1011 structured sources using graph-based label propagation algorithm . ( Talukdar and Pereira
D09-1102 Kernel-based Similarity The key issue in label propagation is how to compute the similarity
D08-1061 a graph-based semi-supervised label propagation algorithm for acquiring opendomain
D11-1011 Shang-Hua Abstract Class-instance label propagation algorithms have been successfully
D09-1102 have systematically evaluated the label propagation algorithm on global learning
D08-1061 Adsorption is a general framework for label propagation , consisting of a few nodes annotated
D09-1102 labeling matrix Y . During the label propagation process , the label distribution
D09-1149 Bootproject ( Zhang , 2004 ) , Label propagation ( Chen et al. , 2006 ) with our
D09-1102 resolution is still unclear . 3 Label Propagation In the LP algorithm ( Zhu and
D08-1061 3 . Demonstrated a graph-based label propagation algorithm that given as little
D09-1149 corpus . Both Bootproject and Label propagation select 100 initial instances
hide detail