Seminar: Antonio LIETO

12 January 2017

LLC Seminar


15.00 – 17.00


Aula di Antica

Palazzo Nuovo (piano II)

via Sant'Ottavio 20, Torino



Antonio LIETO (Università di Torino)


Levels of representation in cognitive architectures: The role of conceptual spaces


Abstract. Cognitive architectures have been historically introduced (i) to capture, at the computational level, the invariant mechanisms of human cognition, and (ii) to reach human level intelligence, also called AGI, by means of the realization of articial artifacts built upon them. During the last decades, many general Cognitive Architectures (CAs) have been realized adopting different assumptions about the representation of their knowledge level. Some of them (e.g. SOAR) adopt a classical symbolic approach, some (e.g. LEABRA) are based on a purely connectionist model, while others (e.g. CLARION) adopt a hybrid approach combining connectionist and symbolic representational levels. Additionally, some attempts trying to extend the representational capacities of CAs by integrating diagrammatical representations and reasoning are also available (e.g. biSOAR). In this talk, I will argue that none of the proposed representational approaches is able, alone, to account for all aspects of cognition involved in knowledge processing tasks and that, in this perspective, the problem arises of their integration in a theoretically and cognitively motivated way. As a consequence of this evidence I will show that the Knowledge Level of the current CAs is affected by several, foundational, problematic aspects. I will finally show how some of these problems can be dealt with by recurring to the framework of Conceptual Spaces (Gärdenfors, 2000) used as a lingua franca for the grounding of different representational layers in a general CA. Some examples showing the benefits of the extension of the Knowledge Level of a CA with Conceptual Spaces will be presented and discussed.


This talk describes joint work with: Antonio Chella, Marcello Frixione, Peter Gärdenfors, Christian Lebiere, Alessandro Oltramari, Daniele P. Radicioni, and Valentina Rho. Papers are available at