摘要
在融媒体平台高度互联的系统架构中,API作为跨模块调用的核心接口,其行为复杂性与动态性显著提升了平台面临的安全风险。本文面向多类型服务场景,构建了基于调用行为图与语义标签融合的行为识别算法,结合上下文感知机制设计了动态权限调度模型,旨在实现更细粒度、更具前瞻性的权限控制与风险防御策略。实验结果表明,该机制在保障访问效率的同时,能有效提升系统的安全弹性与可控性。
In the highly interconnected system architecture of integrated media platforms,API,as the core interface for cross module calls,significantly increases the security risks faced by the platform due to its complex and dynamic behavior.This article focuses on multi type service scenarios and constructs a behavior recognition algorithm based on the fusion of call behavior graph and semantic tags.It combines context aware mechanism to design a dynamic permission scheduling model,aiming to achieve finer grained and more forwardlooking permission control and risk prevention strategies.The experimental results show that this mechanism can effectively improve the security resilience and controllability of the system while ensuring access efficiency.
作者
黄金光
HUANG Jinguang(Maoming Daily,Maoming Guangdong 525000)
出处
《软件》
2025年第11期155-157,共3页
Software
关键词
融媒体平台
API调用
行为识别
动态权限
integrated media platform
API call
Behavior recognition
Dynamic permissions
