D15-1269 Table 2 shows the accuracy rate of concept selection . Here , we excluded the functional
D15-1269 used . We compared the results of concept selection with handlabeled ones . Table
W14-1107 belongs to the concept . Method for Concept Selection Different types of frames are
D15-1269 the baseline . To evaluate the concept selection of words , 98 words in teaching
W14-1107 sub-trees in original ontologies . The concept selection tool suggests the appropriate
W09-1325 to a sentence and depend on the concept selection . They take boolean values or
D15-1269 proposed method . It is clear that concept selection is improved by using the BHMM
E91-1046 , however , is not confined to concept selection as in current knowledge-based
E87-1032 take is not yet known . After concept selection , the local coherence operators
D15-1269 learning phase , the MI results of concept selection for each word are used as the
P09-1070 concepts describing the data . The concept selection smoothing parameter is set as
C04-1083 database . For the purpose of concept selection , only the first 1000 documents
W09-1802 update the length constraint and concept selection accordingly . Figure 3 gives
D13-1156 imposed by the relation between concept selection and sentence selection : selecting
W15-4722 ) visual cues are helpful for concept selection , although the precision is reduced
W15-4722 woman followed by car or boot . Concept selection is performed in a greedy fashion
E87-1032 by Cullingford ( 1986 ) . The concept selection cycle builds a single ERKS meaning
W14-5421 Methodology for predefined visual concept selection , and ( 6 ) Applying ML to extract
W15-4722 i.e. bi - grams ) is helpful for concept selection ; ( ii ) frequency of concepts
hide detail