P08-1079 2006 ) , we do not perform any pattern ranking . Instead we produce ( pos -
P06-1040 for comparison ; they are not pattern ranking algorithms . The best method
P06-1040 better than several alternative pattern ranking algorithms , based on tf-idf
W06-2207 investigate several algorithms for pattern ranking , the most important component
P06-1040 for their answers . All of the pattern ranking algorithms are given exactly
W06-2919 negative information was used during pattern ranking in this experiment . This suggests
N07-1017 more prone to low precision . Pattern Ranking and Selection . Espresso ranks
P03-1044 than a binary partition . 3 . Pattern Ranking : Every pattern appearing in
P06-1040 problem , compared to various other pattern ranking methods . The bottom two rows
P08-1027 , 2006 ) and for unsupervised pattern ranking ( Turney , 2006 ) . Davidov et
P07-1074 applying a similarity measure for pattern ranking , a fair comparison is not possible
P07-1074 describes our experiments with pattern ranking , filtering and rule induction
W09-1308 discovery system described above . 2.4 Pattern Ranking A newly discovered pattern is
W13-1015 Such figures are the basis for pattern ranking and are used in the repository
W11-2165 law play role political economic Pattern Ranking Statistics Several choices exist
D11-1076 Espresso performs pattern induction , pattern ranking and selection using previously
W09-1308 subjects with asthma " . But our pattern ranking algorithm will assign a low score
W04-1206 filter . 2.5 Permanent extraction pattern ranking . After the filtering of the
W09-1308 to very similar results . 3.3 Pattern Ranking Similar to the results observed
hide detail