Research Article
NN-QuPiD Attack: Neural Network-Based Privacy Quantification Model for Private Information Retrieval Protocols
Table 5
Performance of NN-QuPiD Attack with RNN algorithm and query string feature vector under different Epochs.
| | Epochs | RNN LSTM | RNN BiLSTM | | Precision | Recall | Precision | Recall |
| | 5 | 0.936 | 0.432 | 0.933 | 0.451 | | 10 | 0.95 | 0.457 | 0.938 | 0.452 | | 15 | 0.936 | 0.462 | 0.945 | 0.468 | | 20 | 0.936 | 0.463 | 0.92 | 0.485 | | 25 | 0.936 | 0.462 | 0.928 | 0.479 | | 30 | 0.934 | 0.467 | 0.945 | 0.493 | | 50 | 0.932 | 0.459 | 0.925 | 0.493 | | 100 | 0.934 | 0.466 | 0.923 | 0.512 | | 150 | 0.932 | 0.47 | 0.931 | 0.51 |
|
|