We present CLARC (Cryptographic Library for Anonymous Reputation and Credentials), an anonymous credentials system (ACS) combined with an anonymous reputation system. Using CLARC, users can receive attribute-based credentials from issuers. They can efficiently prove that their credentials satisfy complex (access) policies in a privacy-preserving way. This implements anonymous access control with complex policies. Furthermore, CLARC is the first ACS that is combined with an anonymous reputation system where users can anonymously rate services. A user who gets access to a service via a credential, also anonymously receives a review token to rate the service. If a user creates more than a single rating, this can be detected by anyone, preventing users from spamming ratings to sway public opinion. To evaluate feasibility of our construction, we present an open-source prototype implementation.

Reputation systems are used to compute and publish reputation scores for services or products. We consider reputation systems where users are allowed to rate products that they purchased previously. To obtain trustworthy reputations, they are allowed to rate these products only once. As long as users rate products once, they stay anonymous. Everybody is able to detect users deviating from the rate-products-only-once policy and the anonymity of such dishonest users can be revoked by a system manager. In this paper we present formal models for such reputation systems and their security. Based on group signatures presented by Boneh, Boyen, and Shacham we design an efficient reputation system that meets all our requirements.

Group signatures, introduced by Chaum and van Heyst [15], are an important primitive in cryptography. In group signature schemes every group member can anonymously sign messages on behalf of the group. In case of disputes a dedicated opening manager is able to trace signatures - he can extract the identity of the producer of a given signature. A formal model for static group signatures schemes and their security is defined by Bellare, Micciancio, and Warinschi [4], the case of dynamic groups is considered by Bellare, Shi, and Zhang [5]. Both models define group signature schemes with a single opening manager. The main difference between these models is that the number of group members in static schemes is fixed, while in dynamic schemes group members can join the group over time.