From Static to Dynamic Access Control Policies via Attribute-Based Category Min...
summary
A miner synthesizes category-based access-control policies from logs and matrices, producing logical category formulas, allowed actions and a proof of termination; it targets dynamic switching when attributes or events change and flags inconsistencies. The algorithm combines NLP-based similarity to group entities with an FP-growth split/merge check to ensure members share identical authorizations, aims to output a minimal set of categories, and reports synthesized policies plus counterexamples when mining fails.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:17Access control challenges
- 3:43Category-based meta model
- 6:14Meta model principles
- 9:28Static vs dynamic policies
- 12:18Mining algorithm details
- 15:51Future work
- 16:16Q&A
related talks
42mDeclarative Access Management Platform: From Roles to Dynamic Policies
30mRedefining Access Control: Scaling Policy as Code for Humans and AI Agents
42mDeclarative access management platform: from roles to dynamic policies
20mFrom Alignment to Access Control: A Unified View of GenAI...
39mAttribute Based Access Control Framework for IoT
24mExpanding beyond authentication with emerging authorization standards
32mAgentic AI Governance: Data Foundations & Real-Time Policy Enforcement
30mYour AI Has Too Much Access: Authorization for Coding Agents