Automated User Experience Research is made possible by Artificial Intelligence

Automated User Experience Research is made possible by Artificial Intelligence

Rahul Sharma,

,

Abstract. This paper describes an approach to a prototypic system that aims to automate user research activities. Building digital products that fulfil user needs requires an analysis of its users, their contexts and their tasks. These user research activities commonly require a large amount of human effort, resulting in collections of unlinked quantitative and qualitative data sets. Our system aims to analyse and combine “human-made” user research data and extract meaningful insights through machine learning methods. This paper highlights the challenges I faced while building the first prototypes, our need for labelled data and how I managed to overcome its scarcity. Finally, I describe the first prototype which consists of a keyword extractor, sentiment extractor and text annotator, providing a programmatic way to analyse user research data.

Keywords: Machine learning · Natural language processing · Neural networks · User research · Qualitative analysis · Quantitative analysis