Table 3. Run time and F-measure of our automated background knowledge selection methodology, and F-measure of the corresponding baseline and optimal alignments.
| Matching Task | Automated Selection | Baseline | Optimal1 | |
| Time (s)2 | F-measure | F-measure | F-measure | |
| Mouse-Human | 120 | 91.1% | 81.2% | 91.1% |
| FMA-NCI small | 140 | 86.1% | 83.6% | 86.1% |
| FMA-SNOMED small | 190 | 78.1% | 76.3% | 78.1% |
| FMA-SNOMED whole | 800 | 76.5% | 75.6% | 76.6% |
| SNOMED-NCI small | 600 | 72.7% | 69.5% | 72.8% |
| SNOMED-NCI whole | 1500 | 70.4% | 68.2% | 70.4% |
1
Alignment obtained with the manually selected combination of background knowledge sources that leads effectively to the highest F-measure.
2
Experiments were run in a desktop computer with an Intel(R) Core(TM) i5-4570 CPU @ 3.20GHz and 16 GB RAM.