Application of Voice Assistant Technology to Teaching Software Architecture Design: A Case Study of Amazon's Alexa

Abstract

Over the last few years, virtual voice assistants have become more widespread in different fields. The purpose of this research is to develop and perform a quality-based evaluation of the model for voice assistant integration in the software engineering classroom to teach software architecture design. The model creation included identification of general software design principles, SOLID software design principles, software design strategies, and software design patterns that consider the software design process from different perspectives and different abstraction levels. As a result, software architecture is created, and the presented concepts can be considered as building blocks of software architecture design. They can be related to the software quality attributes that refer to the non-functional requirements. The design of skills and intents of a voice assistant as well as the creation of a user interaction model was considered through the Amazon Alexa platform. For this purpose, different teaching and learning resources including textbooks, videos, presentations, and source code, were used. Furthermore, in order to investigate possibilities of applying the model in the software design classroom, an experiment was conducted. The experiment included identification of software system architectures as well as their evaluation in the context of software quality metrics. Despite the experiment being limited, preliminary results indicated that the design of the model for voice assistant integration is feasible. Based on the completed evaluation, it was concluded that the implementation of the proposed software design model enables application of generic, general, and stable software solutions that consequently facilitates the maintenance and evolution of software.