| D15-1163 | accuracy and cover - age . We can do | graph pruning | simply by choosing to use different |
| D12-1076 | decided by SCAN . The semantic | graph pruning | threshold is set to 0.27 tuned |
| W07-0721 | experiment with several ways of | graph pruning | . Addition - ally , for each |
| N10-1108 | redundant patterns are removed via a | graph pruning | algorithm . In experiments on |
| W07-0721 | n-gram length greater than 5 . 5.4 | Graph pruning | The more complex is the reordering |
| W07-0721 | , we experiment with different | graph pruning | which guarantees the translation |
| N12-1051 | task-specific algorithms such as | graph pruning | , edge weighting , and so on |
| J13-3007 | taxon - omy , we perform a step of | graph pruning | , as described in the next section |
| W07-0721 | direc - tions ) . It is shown that | graph pruning | guarantees the efficiency of |
| S14-2034 | pruning and packing algorithms . 3.1 | Graph pruning | Our PRUNING algorithm removes |
| D14-1034 | where I stands for a preposition . | Graph Pruning | . The edge set of our model consists |
| W07-0721 | efficiency : we analyze different | graph pruning | and we show the very low increase |
| W05-0834 | systems are phrase-based . + Their | graph pruning | method is suboptimal as it considers |
| J13-3007 | neither domain heuristics nor the | graph pruning | could completely eliminate the |
| S15-2155 | between individual node pairs and | graph pruning | or edge collapsing ( Kozareva |
| W07-0721 | have experimented with different | graph pruning | showing that best translation |
| S10-1094 | . 4 Knowledge-Based WSD using | Graph Pruning | Wordnet can be viewed as a graph |
| D14-1088 | described in Section 2.2.4 . Step 3 : | Graph pruning | ( line 18 ) The hypernym graph |
| J13-3007 | regard to the two versions of our | graph pruning | algorithm , we found that TREE |
| J13-3007 | heuristic rules , we devised a novel | graph pruning | algorithm , based on the Chu-Liu/Edmonds |
