This paper presents the Temporal Abstract Behavior Representation Model (TABRM). Current techniques for representing behaviors suffer from a lack of abstract representation capability and do not possess the robustness to be used in diverse environments. Without abstraction, the representation of behaviors becomes computationally complex due to the wealth of detail required to enumerate all attributes associated with the environment and the potential courses of action. As a result existing behavior representations tend to be restricted to a limited number of environments or behaviors. TABRM addresses these limitations by using abstraction to define a small number of abstract behaviors and environments. Through the use of abstracted behaviors and environments, TABRM is able to operate using a small decision tree to determine the most appropriate behavior for a given environment. TABRM translates detailed environments into an abstract representation, determines an appropriate abstract behavior, and translates the results back to the detailed environment to produce suitable actions. Decision making within the abstract realm allows an appropriate behavior to be selected regardless of the actual detailed environment. This provides robustness in the model, which is demonstrated within this paper through the application of the model to a range of behavioral domains.
| Identifer | oai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:honorstheses1990-2015-1675 |
| Date | 01 January 2007 |
| Creators | Mansfield, Rachel |
| Publisher | STARS |
| Source Sets | University of Central Florida |
| Language | English |
| Detected Language | English |
| Type | text |
| Source | HIM 1990-2015 |
Page generated in 0.0022 seconds