Abstraction Level Regulation of Cognitive Processing Through Emotion-Based Attention Mechanisms

ion Level Regulation of Cognitive Processing Through Emotion-Based Attention Mechanisms Luís Morgado1,2 and Graça Gaspar2 1 Instituto Superior de Engenharia de Lisboa Rua Conselheiro Emídio Navarro, 1949-014 Lisboa, Portugal [email protected] 2 Faculdade de Ciências da Universidade de Lisboa Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugal [email protected] Abstract. In domains where time and resources are limited, the ability to balance resource consumption according to the problem characteristics and to the required solution quality is a crucial aspect of intelligent behavior. Growing evidence indicates that emotional phenomena may play an important role in that balance. To support this view we propose an agent model where emotion and reasoning are conceived as two symbiotically integrated aspects of cognitive processing. In this paper we concretize this view by extending emotionbased regulation of cognitive activity to enable an active control of the abstraction level at which cognitive processes operate through emotion-based attention mechanisms, thus allowing a dynamical adjustment of the resources used. Experimental results are presented to illustrate the proposed approach and to evaluate its effectiveness in a scenario where reasoning under time-limited conditions in a dynamic environment is required. In domains where time and resources are limited, the ability to balance resource consumption according to the problem characteristics and to the required solution quality is a crucial aspect of intelligent behavior. Growing evidence indicates that emotional phenomena may play an important role in that balance. To support this view we propose an agent model where emotion and reasoning are conceived as two symbiotically integrated aspects of cognitive processing. In this paper we concretize this view by extending emotionbased regulation of cognitive activity to enable an active control of the abstraction level at which cognitive processes operate through emotion-based attention mechanisms, thus allowing a dynamical adjustment of the resources used. Experimental results are presented to illustrate the proposed approach and to evaluate its effectiveness in a scenario where reasoning under time-limited conditions in a dynamic environment is required.