Provenance Information-based Trust Evaluation using Cooperation Pattern for Self-Adaptive Systems

Trust is one of the major factors that a system needs to achieve the secure interaction. Trust can represent the social and reliability perspective of the security. This characteristic makes trust the soft security factor. There are some attempts to evaluate trust, however, they do not consider the characteristics of self-adaptive systems such as self-* properties, openness, and uncertainty. In this paper, we propose provenance information-based trust evaluation using the cooperation pattern for self-adaptive systems. Provenance information contains historical records of the system. By using provenance information, we can evaluate a system's trust value without any help from outside. Even though the system has a good trust value, the value can be manipulated by other systems which have the malicious intention. To identify the malicious intention, we propose a cooperation pattern that represents the singularity of provenance information and gives incentives or penalties to the trust value.