D09-1018 supervised framework , we use Iterative Collective Classification ( ICA ) , which facilitates machine
D12-1053 sentences were considered for collective classification . The cost value was set to 500
H05-1056 Further , it is not clear if a collective classification approach can scale to modeling
H05-1043 subtasks is cast as an unsupervised collective classification problem and solved using the
P11-1151 outputs of machine learners into collective classification algorithms . Our experimental
N12-1093 which they view as an instance of collective classification ( Barzilay and La - pata , 2005
N06-1027 discussion threads using similar collective classification ap - proaches . We next discuss
P11-1151 with network structures , where collective classification can yield better performance
P11-1151 Congressional floor - debate transcripts . Collective classification techniques are used to take advantage
D09-1018 paradigms are used : a supervised collective classification framework and an unsupervised
P10-2054 ting , the role-filling task is a collective classification problem , since we know each
D09-1018 and Jensen , 2000 ) , which is a collective classification algorithm shown to perform consistently
D13-1077 Translate . <title> Cascading Collective Classification for Bridging Anaphora Using a
P11-1151 provide a formal definition of collective classification and describe the ConVote corpus
E14-1040 structure , we utilize an approximate collective classification algorithm , loopy belief propagation
N06-1027 describe a dependency-network based collective classification method to classify email speech
P11-1151 approaches based on three alternative collective classification algorithms ( Section 3 ) . We
P11-1151 from its observed attributes . Collective classification whether a given pair of instances
D12-1082 2010 ) studied hierarchical and collective classification using 92 types , and FIGER (
P11-1151 sentiment classification . <title> Collective Classification of Congressional Floor-Debate
hide detail