Explainable AI

While eXplainable AI (XAI) is only recently gaining widespread visibility, the Machine Learning (ML), Artificial Intelligence (AI) and Recommender Systems literature contain a long history of work on explanations. The distinction was made early on between transparency that explains the inner logic of a model and justification that decouples the explanation from the model. The later category is also researched under the umbrella of model-agnostic approaches. The overall objective of this project is to enable non-expert users to understand the outcome of ML models as a basis to support informed decision-making. As individual users have different needs, the long-term goal of our research is to adapt explanation interfaces to the specific needs, characteristics, and context of each user.