Table 6 The CTI event arguments extraction accuracy ranking by proposed method (CAFIIE) in four datasets.
From: Cyberattack event and arguments extraction based on feature interaction and few-shot learning
Entity Type | CASIE | Entity Type | DNRTI | ||||
|---|---|---|---|---|---|---|---|
P | R | F1 | P | R | F1 | ||
Money | 0.9431 | 0.9316 | 0.9373 | Loophole | 0.9861 | 1 | 0.9930 |
CVE | 0.8694 | 0.8971 | 0.8830 | Sample-file | 0.9645 | 0.8589 | 0.9086 |
Patch | 0.8168 | 0.8136 | 0.8561 | Features | 0.9173 | 0.9914 | 0.9529 |
Device | 0.8079 | 0.8463 | 0.8267 | Industry | 0.9126 | 0.9457 | 0.9289 |
Organization | 0.7423 | 0.7842 | 0.7627 | Security-team | 0.8974 | 0.8487 | 0.8724 |
Entity Type | MalwareTextDB | Entity Type | Private Dataset | ||||
|---|---|---|---|---|---|---|---|
P | R | F1 | P | R | F1 | ||
Action | 0.5994 | 0.6823 | 0.6382 | Property | 0.6876 | 0.6371 | 0.6614 |
Relation | 0.6512 | 0.6470 | 0.6491 | ||||
Entity | 0.5981 | 0.6154 | 0.6066 | Instance | 0.6145 | 0.6374 | 0.6257 |
CWE | 0.6108 | 0.6462 | 0.6280 | ||||
Modifier | 0.5547 | 0.5961 | 0.5747 | Software | 0.6069 | 0.6314 | 0.6189 |
