Translating text to events happening in a 3D virtual world is one way of evaluating machine understanding of human language. This entails several challenges including how to represent sentences and full discourses that are grounded in the physical and social world. Solving these challenges inspired by how the human brain processes language is an important goal of the ERC Advanced Grant CALCULUS. When grounding language, one interesting question is how the structure of language and the structure of the perceived world interact, which leads to representations that naturally integrate structure and which might be inspiring when representing space, time and causality. In human language understanding commonsense knowledge plays an important role. Especially how to acquire, represent and fuse commonsense knowledge in the understanding process are pertinent research questions. Another problem regards the representation of a memory that stores such knowledge as well as content that occurred previously in a discourse. Finally, intelligent models of human language understanding continually learn to improve their skills, which often entails learning from data which are not independent and identically distributed.
Marie-Francine (Sien) Moens is full professor at the Department of Computer Science at KU Leuven, Belgium. She is the director of the Language Intelligence and Information Retrieval (LIIR) research lab and a member of the Human Computer Interaction group. Her main direction of research is the development of novel methods for automated content recognition in text and multimedia using machine learning and exploiting insights from linguistic and cognitive theories.