Trustworthiness and Quality of Context Information

Abstract

Context-aware service platforms use context information to customize their services to the current users' situation. Due to technical limitations in sensors and context reasoning algorithms, context information does not always represent accurately the reality, and Quality of Context (QoC) models have been proposed to quantify this inaccuracy. The problems we have identified with existing QoC models is that they do not follow a standard terminology and none of them clearly differentiate quality attributes related to instances of context information (e.g. accuracy and precision) from trustworthiness, which is a quality attribute related to the context information provider. In this paper we propose a QoC model and management architecture that supports the management of QoC trustworthiness and also contributes to the terminology alignment of existing QoC models. In our QoC model, trustworthiness is a measurement of the reliability of a context information provider to provide context information about a specific entity according to a certain quality level. This trustworthiness value is used in our QoC management architecture to support context-aware service providers in the selection of trustworthy context providers. As a proof of concept to demonstrate the feasibility of our work we show a prototype implementation of our QoC model and management architecture.