Monday, 1 April, 2019 - 13:30

Actionability in Cognitive Robotics

Sergei Nirenburg and Marjorie McShane (Rensselaer Polytechnic Institute, Troy, NY, USA)

Ever wondered why people interrupt each other? It is not just because they are impolite. Often this is because hearers understand what speakers want to say before the end of the utterance. That is, the meaning of the utterance becomes actionable before the speaker has finished.

Actionability-based language understanding shifts the overall objective of computational linguistics from deriving complete semantic and discourse/pragmatic interpretations of language inputs to understanding just enough to confidently maintain a conversation, learn from a text, or participate in a team task.

In this talk, we will motivate and illustrate the methodological shift toward actionability by considering the desiderata for developing artificial intelligent agents that are human-like in their abilities to understand, learn, and explain.


Dr. Sergei Nirenburg is Professor of Cognitive Science, Professor of Computer Science and Head of Department of Cognitive Science at Rensselaer Polytechnic Institute. He is Co-Director, with Dr. McShane, of Rensselaer’s Language-Endowed Intelligent Agents (LEIA) Lab. Dr. Nirenburg has worked in the areas of cognitive science, artificial intelligence and natural language processing for over 35 years, leading R&D teams of up to 80. He has written two and edited six books and published over 200 scholarly articles in journals and peer-reviewed conference proceedings. Dr. Nirenburg has a PhD in Linguistics from The Hebrew University of Jerusalem, Israel. Before coming to Rensselaer, he taught in the Computer Science Departments at the Hebrew University, Colgate University, Carnegie Mellon University, New Mexico State University and University of Maryland Baltimore County.

Dr. Marjorie McShane is Associate Professor of Cognitive Science and Co-Director of the Language-Endowed Intelligent Agents (LEIA) lab at Rensselaer Polytechnic Institute. She earned her PhD in Slavic Languages and Lingusitics from Princeton University and had prior research appointments at New Mexico State University and the University of Maryland Baltimore County. She has published two books, 40 articles in journals and books, and over 50 papers in refereed conference proceedings. Over the past 20 years, McShane has headed up development efforts on deep language understanding, knowledge acquisition from human experts for use in agent systems, and the development of agent-oriented cognitive models from that expert knowledge. McShane is particularly interested in the extent to which the interpretation of language is reliant on extra-linguistic knowledge, and the process of distilling expert knowledge into the knowledge necessary to support interactive computer simulations.