LinkedIn如何将用户限制系统扩展到每秒500万次查询

Disclaimer: The details in this post have been derived from the LinkedIn Engineering Blog. All credit for the technical details goes to the LinkedIn engineering team. The links to the original articles are present in the references section at the end of the post. We’ve attempted to analyze the details and provide our input about them. If you find any inaccuracies or omissions, please leave a comment, and we will do our best to fix them.

免责声明：本文中的细节来源于LinkedIn工程博客。所有技术细节的归功于LinkedIn工程团队。原始文章的链接在文章末尾的参考部分中列出。我们尝试分析这些细节并提供我们的见解。如果您发现任何不准确或遗漏的地方，请留言，我们将尽力修正。

One of the primary goals of LinkedIn is to provide a safe and professional environment for its members.

LinkedIn的主要目标之一是为其成员提供一个安全和专业的环境。

At the heart of this effort lies a system called CASAL.

这一努力的核心是一个名为 CASAL 的系统。

CASAL stands for Community Abuse and Safety Application Layer. This platform is the first line of defense against bad actors and adversarial attacks. It combines technology and human expertise to identify and prevent harmful activities.

CASAL代表社区滥用和安全应用层。该平台是抵御恶意行为者和对抗性攻击的第一道防线。它结合了技术和人类专业知识，以识别和防止有害活动。

The various aspects of this system are as follows:

该系统的各个方面如下：

• ML Models: The ML models analyze patterns in user behavior to detect anything unusual or suspicious. For example, if a user suddenly sends hundreds of connection requests to strangers or repeatedly posts harmful content, the system can flag these activities for review.

  机器学习模型：机器学习模型分析用户行为中的模式，以检测任何异常或可疑的行为。例如，如果用户突然向陌生人发送数百个连接请求或反复发布有害内容，系统可以将这些活动标记为待审查。

• Rule-Based Systems: These systems work based on pre-defined rules. Think of them as guidelines that help the platform decide what’s acceptable and what’s not. For instance, certain words or actions that violate LinkedIn’s policies (like hate speech or spam) automatically trigger alerts.

  基于规则的系统: 这些系统基于预定义的规则工作。可以将它们视为帮助平台决定什么是可接受的，什么是不可接受的指南。例如，某些违反 LinkedIn 政策的词语或行为（如仇恨言论或垃圾邮件）会自动触发警报。

• Human Review Processes: Not everything can be left to machines. A dedicated team of human experts steps in to review flagged activities and make decisio...