A learning model for trustworthiness of context-awareness services

When ubiquitous computing devices access a context-awareness service, such as a location service, they need some assurance that the quality of the information received is trustworthy. However, the trustworthiness of a service cannot be determined by the service itself but must be decided externally to the service. Furthermore, the trustworthiness of a service provider may be dynamic, depending on current environmental conditions. We propose a learning model that uses binary positive/negative feedback from service consumers and cross-validation with other service providers to adjust the dynamic trustworthiness of a service provider.