An IoT Architecture for Personalized Recommendations over Big Data Oriented Applications

The paper presents an innovative Internet of Things architecture for building personalized services in the smart city context. The main blocks of the presented implementation comprise data flows implemented through Node-Red, Neo4j data store for handling the smart city big data and a recommendation service which is applied in order to offer personalized recommendations to the users. The current work studies integration of the various components, the modelling approach for user generated data combined with open big data and proceeds with the appropriate reference implementation and experimentation to validate the personalized recommendation services for innovative citizen-centric applications and use cases. Moreover, we study and validate performance issues of this Neo4j based recommendation service and evaluate it as a useful appliance for real-time big data application.