Information Collection Platform for Smart Nudging. A Microservice-Based Approach.

Abstract

This thesis aims to explore the problem of integrating heterogeneous data sources into the Smart Nudge system. The Smart Nudge system is a system that produces personalised nudges that are contextually relevant to each user. The system relies on access to live data that could be constructed and presented in specific ways to influence users behaviour towards an agreed-upon goal. The goal is to ascertain the suitability of a microservice-based approach to designing the component that is responsible for integrating various data sources. A small prototype of two microservices provided a practical look at integrating real-world sources, namely a Norwegian weather service and a bus tracking service in Chicago. The proposed architecture is analysed using a set of requirements derived from a theoretical examination of the Smart Nudge system and a general theoretical look at decomposition techniques used to evaluate microservice architectures. Evaluating the prototype revealed that the Smart Nudge system is highly dependant on augmenting data sources with additional meta-data to produce personalised nudges. The analysis indicates that a data-driven microservice-based architecture seems well suited to resolving some of the problems and requirements that are somewhat unique to the Smart Nudge system setting.