Probabilistic Agents in Deterministic Audits: Evaluating Multi-Agent Systems for Automated Audits Based on the German IT-Grundschutz
- 类型:arxiv
- 标识:2606.25622
- 链接:http://arxiv.org/abs/2606.25622v1
- 主分类:agent
- 形态:benchmark
- 被引:1
- 被引来源:Semantic Scholar
- S2被引:1
- OpenAlex被引:0
- 影响力被引:0
- TLDR:This paper presents the technical implementation and empirical evaluation of a Multi-Agent System (MAS) architecture combined with Hybrid Retrieval Augmented Generation (HybridRAG) for the partial automation of IT-GS certification and introduces two novel technical contributions to the MAS architecture to enforce the compliance rigor.
- OpenAlex ID:W7165941852
- OpenAlex DOI:10.48550/arxiv.2606.25622
- DOI:10.48550/arxiv.2606.25622
- DOI来源:OpenAlex
- 开放获取:green
- 开放获取链接:https://doi.org/10.48550/arxiv.2606.25622
- OpenAlex更新:2026-07-19
- 副分类:evaluation
- 待LLM分类:否
- 标题中文:确定性审计中的概率 Agent:基于德国 IT-Grundschutz 的多 Agent 系统自动化审计评估
- TLDR中文:本文展示了用于 IT-Grundschutz(IT-GS)认证部分自动化的多 Agent 系统(MAS)架构结合混合检索增强生成(HybridRAG)的技术实现与实证评估,并为强化合规严谨性引入两项新的 MAS 架构技术贡献。
- 来源文件:
- /inbox/tom/_candidates/2026-06-25-agent-rag-longcontext-candidates.json
- [S2 enrich]
- [OpenAlex backfill]