SLICE-based Trustworthiness Analysis system

We introduce system architecture and its functional blocks for the trustworthiness analysis in which the IoT data mining algorithm is used in order to provide trust based services in the field of the connected car. Our system trains dangerous driving behavior(DDB) detection model from the OBD(On Board Diagnosis) data and predicts whether a driver drives dangerously. For this, the system was developed based on the Self Learnable IoT Common SW Engine(SLICE) enabling context awareness, inference and learning on the IoT node by itself. More specifically, while the existing approach predicts the dangerous driving depending on the velocity of the car, the proposed system finds the relevant OBD parameters except for the velocity and makes the decision tree for prediction.