Table 1 Comparison of different implementations and optimizations. The table shows actual times spent with updating and querying the generalization lattice with different implementations as well as improvements in performance compared to closest less sophisticated implementation option
From: Efficient and effective pruning strategies for health data de-identification
| Â | Option 1 | Option 2 | Option 3 | Option 4 |
|---|---|---|---|---|
Dataset | Simple list | List with antichain | Prefix tree with antichain | Optimized tree with antichain |
Adult | 0.247s | 0.230s (6.88 %) | 0.061s (73.48 %) | 0.055s (9.84 %) |
Cup | 0.658s | 0.525s (20.21 %) | 0.232s (55.81 %) | 0.218s (6.03 %) |
Fars | 0.364s | 0.240s (34.07 %) | 0.074s (69.17 %) | 0.071s (4.05 %) |
Atus | 0.300s | 0.185s (38.33 %) | 0.095s (48.65 %) | 0.081s (14.74 %) |
IHIS | 0.437s | 0.251s (42.56 %) | 0.124s (50.60 %) | 0.093s (25.00 %) |
SS13ACS | 21.730s | 15.032s (30.82 %) | 1.861s (87.62 %) | 1.797s (3.44 %) |