Evaluation on manual labeled testing set

Here, we list the indicators helping us to manually label 25 SR-CRs as a supplement to demonstrate the diversity of SR-CRs. Notice that, using only these indicators to figure out SR-CRs will bring in plenty of false positives, as they just provide a clue showing the potential of an SR-CR.

id-1: expert's knowledge;

id-2: expert's knowledge;

id-3: expert's knowledge;

id-4: ``... ciphering may fail ...'';

id-5: ``... at risk of eavesdropping ...'';

id-6: ``... security threat ...'';

id-7: ``... session keys ... compromised ... if ... with no encryption ...'';

id-8: ``... attacker...'';

id-9: ``... authentication ... will fail ...'';

id-10: ``... sensible information may be revealed ...'';

id-11: ``... skipping of authentication for malicious UE ...'';

id-12: ``... undermine security guarantees ...'';

id-13: ``... integrity protection ... encryption algorithm ...'';

id-14: ``... leakage of access token ...'';

id-15: ``... DDOS attack ...'';

id-16: ``... fatal security problem remains ...'';

id-17: expert's knowledge;

id-18: ``... security key ...'';

id-19: ``... malicious ...'';

id-20: ``... rogue UE ...'';

id-21: expert's knowledge;

id-22: ``... violation ... user privacy ...'';

id-23: ``... security aspects are not aligned ...'';

id-24: expert's knowledge;

id-25: expert's knowledge;