Ontology summit 2019 communiqué: Explanations

With the increasing amount of software devoted to industrial automation and process control, it is becoming more important than ever for systems to be able to explain their behavior. In some domains, such as financial services, explainability is mandated by law. In spite of this, explanation today is largely handled in an unsystematic manner, if it is handled at all. The decisions of modern AI systems, such as those built on deep neural networks, are especially difficult to explain. The goal of the recent Ontology Summit 2019 was concerned with the role of ontologies for explaining the functioning of a system. More specifically, the Ontology Summit focused on critical explanation gaps and the role of ontologies for dealing with these gaps. The sessions examined current technologies and real needs driven by risks and requirements to meet legal or other standards. The sessions covered explainable artificial intelligence, commonsense reasoning and knowledge, the role of narrative, and explanations in the fields of finance and medicine. The goal of this Communiqué is to foster research and development of approaches to explanations and to drive towards explanation support which can be incorporated into both knowledge engineering processes and ontology design best practices.