Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


osu1198526352.pdf (762.17 KB)

ETD Abstract Container

Abstract Header

Design and use of a bimodal cognitive architecture for diagrammatic reasoning and cognitive modeling

Kurup, Unmesh
http://rave.ohiolink.edu/etdc/view?acc_num=osu1198526352

Abstract Details

2008, Doctor of Philosophy, Ohio State University, Computer and Information Science.
In 1948, Tolman introduced the term cognitive map to refer to the representation of large-scale space in rats that allow them to remember and navigate back to food sources in a maze. Current proposals for the structure of the cognitive map are usually self-contained theories, each with its own set of representations and processes, whereas in reality, the cognitive map and associated processes are part of the larger cognitive apparatus. Thus, while spatial representations and processes may have unique aspects, it is likely that they also share underlying representations and mechanisms with those involved in general problem solving. It is important then, both for reasons of accuracy in cognitive modeling and economy in agent building, that spatial representation and reasoning systems be well integrated with general-purpose problem solvers. However, given the differences in mechanisms, assumptions and constraints among these systems, integrating them is a challenging task. An ongoing area of research in AI and cognitive science has been the unification of cognitive theories within the framework of a cognitive architecture. This work proposes the use of the cognitive architecture methodology to investigate issues in the representation of and reasoning about large-scale space. In particular, we use biSoar, a version of the cognitive architecture Soar that we have augmented with the Diagrammatic Representation System (DRS). Soar is a well known cognitive architecture for constructing general cognitive systems while DRS allows the representation of a collection of diagrammatic objects each with their complete spatiality. In biSoar, information can thus be represented both symbolically and diagrammatically, as appropriate. We exercise biSoar’s capabilities both in problem solving and in cognitive modeling. For problem solving, we use examples from the blocks world domain to show how a bimodal architecture can be advantageous. These examples also serve to demonstrate the bimodal problem solving process. For cognitive modeling, we build biSoar models of three different spatial phenomena. These modeling examples showcase biSoar’s flexibility and versatility as well as its usefulness as a tool in cognitive modeling.
Balakrishnan Chandrasekaran (Advisor)
Computer Science
Diagrammatic Reasoning; Cognitive Architectures; Multi-modal Architectures; Cognitive Modeling; Spatial Reasoning; Frame Problem

Recommended Citations

Citations

  • Kurup, U. (2008). Design and use of a bimodal cognitive architecture for diagrammatic reasoning and cognitive modeling [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1198526352

    APA Style (7th edition)

  • Kurup, Unmesh. Design and use of a bimodal cognitive architecture for diagrammatic reasoning and cognitive modeling. 2008. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1198526352.

    MLA Style (8th edition)

  • Kurup, Unmesh. "Design and use of a bimodal cognitive architecture for diagrammatic reasoning and cognitive modeling." Doctoral dissertation, Ohio State University, 2008. http://rave.ohiolink.edu/etdc/view?acc_num=osu1198526352

    Chicago Manual of Style (17th edition)

osu1198526352
792
© 2007, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.