The research is conducted for and on behalf on Info Support, but its findings are generic and apply to all CUIs in general. CUIs are a new, trending technology with little to no research currently available and new frameworks are necessary for proper implementations of the system. During this research, the Conversation Design Framework (CDF) is created as the solution for specifying conversations of the CUI. The framework consists of seven concrete steps and aims to set a best practice design guideline for specifying an effective and efficient conversation.
This research aimed to the best way of specifying and testing conversational user interfaces in an Agile environment.
After elaborating the CDF, methods are proposed for testing the specifications that are created when using this model. A method of testing the personality is absent as only indirect testing suggestions are given, while many other testing can potentially be automated. Afterward, it is researched how well the well the Agile application of the CDF is and it is concluded that the framework can be implemented with high agility. The findings of this research provide a sound answer to the research question and are a good introduction to research into conversational user interfaces. Further research could be undertaken to create a model of testing chatbot personalities, validating the CDF by applying it to a real system through a case study and creating an evaluation method for bot frameworks.