21 February 2022, Monday. 3PM. London Time. Online over Zoom
Title Representing Knowledge and Thought: Insights from the Design of Radical Representational Systems
Speaker Peter Cheng, University of Sussex
Abstract It is well established in Cognitive Science that the representational systems we use on problems substantially impacts our ease of finding solutions. To maximise the theoretical and empirical leverage available for the study of representational systems, I design novel notations for conceptually challenging topics and novel graphical user-interfaces for information intensive problem-solving. Examples in this talk will highlight new graphical notations for syllogistic and propositional logic. From the task analytic and experimental contrast of such novel representations with extant conventional representations, various insights have been gleaned. STEM subjects should be easy to learn. In representational terms, conventional notations that encode STEM knowledge are typically incoherent. In contrast, effective representational systems possess semantic transparency and are syntactically plastic. Such representations promise a factor of two performance improvement for problem solving and learning. To round off the talk, I will outline a current project that is attempting to build an intelligent tool to automatically select effective representations for different individuals on specific problems.
Bio Peter Cheng is Professor of Cognitive Science and Director of the Representational Systems Lab in the Department of Informatics at the University of Sussex. His bachelor’s degree was in Mechanical Engineering, which was followed by a MA in Philosophy. Then after a PhD in AI at the Open University, Peter held a SERC postdoctoral fellowship at Carnegie Mellon University, where he collaborated with Herbert Simon building computation models of the role diagrams in scientific discovery. His current areas of research include: the cognitive science of representational systems; cognitive learning analytics, and the cognitive science of tactile graphics. Peter’s current projects include: Automating Representation Choice for AI Tools (funded by the EPSRC), and Cybersecurity visualization: collaborative graphical tools for security policies (funded by the Qatar National Research Fund).
Meeting ID: 668 413 8396