A common problem of AI solutions is the lack of trust of potential users due to the often untransparent processes and calculations of the AI system. Users therefore require guidance in understanding, adapting and applying these AI-based solutions. In cooperation with the BASF department for AI solutions, we aim to tackle these problems by developing a conversational user interface that (1) guides users during the interaction with the AI solutions, (2) provides explanations to help users understand the AI solutions, and (3) collects user feedback to further improve the AI solutions. The conversational UI will be developed for BASF forecasting models that are applied to forecast demands, prices and specific KPIs. The overall goal of this project is to understand how conversational user interfaces can be applied in use cases like this to increase trust of users in AI solutions while reducing the workload of the developers for handling user questions and feedback.
If you are interested in working on a Bachelor's or Master's Thesis on conversational user interfaces for feedback collection and user guidance in the context of forecasting AI solutions, please contact Saskia Haug.