Table 3 Results of the evaluation of Algorithm 2 over the SEALINC Media dataset
From: Efficient semi-automated assessment of annotations trustworthiness
# Tags per | % Training | Â | Â | Â | Â | Time |
|---|---|---|---|---|---|---|
reputation | set covered | Accuracy | Precision | Recall | F-measure | (sec.) |
clustered results (cut = 0.6) | ||||||
5 | 8% | 0.73 | 0.88 | 0.81 | 0.84 | 43 |
10 | 19% | 0.82 | 0.87 | 0.93 | 0.90 | 24 |
15 | 31% | 0.83 | 0.87 | 0.95 | 0.91 | 14 |
20 | 41% | 0.84 | 0.87 | 0.96 | 0.91 | 18 |
clustered results (cut = 0.3) | ||||||
5 | 8% | 0.78 | 0.88 | 0.88 | 0.88 | 43 |
10 | 19% | 0.82 | 0.87 | 0.93 | 0.90 | 14 |
15 | 31% | 0.84 | 0.87 | 0.95 | 0.91 | 16 |
20 | 41% | 0.84 | 0.87 | 0.96 | 0.92 | 21 |