P98-2228 taggers we implemented a simple voting system . By comparing the results obtained
N01-1025 representations . For the weighted voting systems , we introduce a new type of
S07-1067 implement is based on a simple voting system . Each classifier returns a score
S13-2022 submitted , one of them considering voting system of the previous four approaches
P13-2127 as underspecified by either the voting system or MACE , but not by both . Table
J01-2002 data sets . Within the simple voting systems , it appears that use of more
E12-2002 environment . The tailor made voting system maximizes the use of the different
N04-1040 and tree models at p .005 . The voting systems did not use any source-pair information
E12-2002 Alliance , 2009 ) . A tailored voting system for multi-label multi-class tasks
E06-1042 successful learner , we introduce a voting system . We use a simple majority-rules
P98-2228 the results obtained from the voting system with those from the decision
S01-1030 especially when it comes to how the voting system should be set up . As the feature
P98-2228 combination of knowledge sources . The voting system provided 59 % correct disambiguation
J11-4006 adjacency model and the unsupervised voting system from Section 6.2.1 . As we described
J01-2002 are clearly better than simple voting systems , at least as long as there is
S10-1053 taken as feature vector , and the voting system predicts a sense . This approach
S10-1053 classifiers and have an arbiter voting system do the final classification step
J01-2002 better accuracy than the simple voting systems on all four data sets . TagPair
P13-2127 Comparison between methods The voting system and MACE provide different sense
N10-1099 Table 1 ) . The baseline majority voting system includes e - rater , GUMS , and
hide detail