Depth learning engine best practices {# concept_1113021 .concept}
Ali cloud Web application firewall using a variety of Web attack detection engine combination to provide comprehensive protection for your site. Web Application Firewall rules engine, way deep learning and semantic analysis engine combination of three kinds, give full play to the powerful Ali cloud intelligence, data analysis and expert system vulnerabilities mining experience advantage. Ali and so on based on attack data analysis crawl 0day vulnerabilities, vulnerability to actively pursue and analyzed by security experts, and finally precipitated as summarized protection rules in the policy. Web application firewall rule policy updates a week, far more than the industry level, to provide users with the fastest, most comprehensive protection.
With the development of the Internet, Web attacks are evolving, the traditional way single means of protection has been unable to meet the escort of complex Internet business requirements, only play the best protective effect through a variety of collaborative detection engine protection order .
Ali Cloud Web Application Firewall rules engine, semantic analysis and deep learning engine a combination of Web attack protection. among them,
- Rules Engine: Alibaba Group over the years relying on offense and defense practice accumulated expertise gathered from protection policy rules.
- Semantic analysis engine: to make up for the traditional rules engine in regular grammar questions about the scope of the weaker features context-free grammar description, rules to reduce security risks and vulnerabilities caused by false positives.
- Depth learning engine: by way of supervised learning, the neural network system relies on powerful algorithms Ali cloud of team building, Ali goes on to attack one hundred million daily data classification training, ultimately unknown risks request by the model in real-time line detection interception, the defense engine to compensate for the lack of other unknown risk 0day vulnerability detection.
Depth learning engine protection case {# section_itw_lsn_0r9 .section}
In general, the rules of regular descriptive rules engine using relatively strong, a strong protective effect on the attack signature request, the best regular rules. And when faced with some of the characteristics of the attack weak requests (eg XSS feature request), even if you enable strict mode Web application attack protection, it may still exist because of potential security risks can not be detected.
For example, you can enable Web application firewall big data engine features deep learning to identify and block attacks weak feature requests the Web application attack protection strict rules unrecognized.
In this case, the request is not XSS attack Web application attack protection rules interception.
Enabling Web Application Firewall Big Data deep learning engine after protection, beyond the regular rules engine detection capabilities of XSS attack request was successfully intercepted.
Meanwhile, a report in Web application attacks can view the detailed log information attack, attack type of deep learning.
