最終更新日:2026-03-25 18:23:34
Advanced Detection is a combined feature set developed based on years of Bot mitigation experience and uses big data analysis and machine learning for dynamic modeling. It helps identify simple Bots, sophisticated Bots, and Advanced Persistent Bots (APBs).
Advanced Detection includes the following capabilities:
Automated Framework Detection: Continuously analyzes requests initiated by automation frameworks, such as browser-like tools, to identify behavior that appears human-like but still follows detectable patterns. This helps distinguish legitimate visitors from Bots that imitate normal user behavior.
High-Risk Sequence Detection: Continuously analyzes script-based Bots that access resources in specific sequences. These malicious requests often follow predictable behavioral paths and access orders that differ from the more random behavior of legitimate users.
AI Detection: Combines years of Bot mitigation experience with historical attack samples to continuously train AI models. This helps identify new attacks and attack variants, and is well suited for scenarios where attack patterns continue to evolve.
Note: This feature uses machine learning algorithms, such as random forests, together with automated labeling for detection. Its optimization goal is to improve Bot recall. Because the model strategy is relatively aggressive in order to cover more potential threats, a higher false positive rate is possible.
In production environments, we recommend placing this feature after other rule engines as a supplementary detection method, and using it together with other detection results for cross-validation to improve overall confidence.
to enter the Security Policy editing page.