Approximating Saml Using Similarity Based Imprecision

With the increasing complexity of networked systems has come the trade-off of security versus functionality; a strictly secured system is often an unusable system. As a consequence, users often entirely bypass security in order to get their job done. We consider how similarity techniques that are used by casebased reasoning systems can be used to provide a degree of control over how strictly/precisely security is enforced. The flexibility to be able to meaningfully control how strictly security is enforced is especially relevant in the emerging Web Services architectures, where a wide variety of different users and heterogeneous systems use a common framework to interoperate with a wide variety of different resources and services. The paper proposes similarity-based imprecision security (SBIS) for the Security Assertion Markup Language (SAML) as an approach to managing security in a web-services environment.