Design Thinking Using Qualitative Data Analysis and Machine Learning

Design Thinking is a human-centered approach that allows continuous feedback by the user through Empathizing, Defining, Testing, Ideating and Prototyping. It mainly focuses on user needs, aspirations, wishes, concerns and frustrations in attempting to solve their problems. The Persona Creation approach follows the process of collecting data from multiple sources including social media platforms or the traditional methods including interviews of different users to cover the different types of behaviors, interactions and goals, questionnaires, or surveys. Condensing gathered data using qualitative data analysis renders assessable domain models that can be shared among and modified by stakeholders, so as to agree on user needs and issues. After agreeing on useful needs and user issues, they are used to generate Personas that represent the different types of users of a specific software product. When both approaches Design Thinking and Persona Creation are incorporated during Agile software development, this would lead to the creation of a successful software product. Successful software products are ones that cover all the needs as mentioned by the product user, also known as user perspectives.

A user perspective refers to the perception of a given user and how they would use the final product. Those are the people who would interact with the software product created and, therefore, the people for whom the software is designed. For this reason, if an application, a website or a functionality that does not meet the final user's needs, this would ultimately result in a failure for the business. Inducing pain points/insights (what is needed by users) should not left totally to the skills of the analyst with little guidance. A systematic and more guidance is needed in this situation. Agile Software Development lack a coherent and explicit technique or open architecture [1] that can accommodate changes mandated by experiments on the ground. In addition, there has to be a method for objectively evaluating resultant prototypes/releases/deliverables at the end of each sprint in a way that can effectively guide path adjustments.

Therefore, in this research, we make use of Design Thinking with software products. Through creating a framework that includes Design Thinking as an elicitation technique. We propose a framework composed of two phases: The first phase is the use of a robust qualitative data analysis method, to achieve models that are rich, and at the same time concise and traceable to their origins. We propose the use of the Grounded Theory method in the analysis and integration of the qualitative data that can characterize user needs, pain points and system requirements, in addition to second layer requirements that are often hard to spot. Second layer requirements are those requirements that are not immediately visible or perceivable by the end-user of a system, or those working with or observing him or her, such as systems and requirements analysts. The source of data for generating grounded theoretical formulations include interviews (of whatever type), observations, online chatter, and documents relating to the immediate and wider contexts of the need phenomenon under study.

The second part of our proposed framework is applying Machine Learning on the data resulting from the first phase so that we are able to automate the Persona creation using Machine learning. Automatic Persona creation via machine learning is used to represent potential users, as an attempt to enhance the requirements of software products since they will necessarily include user perspectives.