Title :
Modelling emotional attachment: An integrative framework for architectures and scenarios
Author :
Dean Petters;Everett Waters
Author_Institution :
School of Social Sciences, Birmingham City University, UK
fDate :
7/1/2015 12:00:00 AM
Abstract :
Humans possess a strong innate predisposition to emotionally attach to familiar people around them who provide physical or emotional security. Attachment Theory describes and explains diverse phenomena related to this predisposition, including: infants using their carers as secure-bases from which to explore, and havens of safety to return to when tired or anxious, the development of attachment patterns over ontogenetic and phylogenetic development, and emotional responses to separation and loss throughout the lifespan. This paper proposes that one way for computational modelling to integrate these phenomena is to organise them within temporally nested scenarios, with moment to moment phenomena organised within ontogenetic and phylogenetic sequences. A number of existing agent-based models and robotic attachment simulations capture attachment behaviour, but individual simulations created with different tools and modelling approaches typically do not integrate easily with each other. Two ways to better integrate attachment model are proposed. First, a number of simulations are described that have been created with the same agent-based modelling toolkit, so showing that moment to moment secure base behaviour and the development of individual differences in attachment security can be simulated with closely related architectural designs. Secondly, an integrative modelling approach is proposed where the evaluation of, and comparison between attachment models is guided by reference to a shared conceptual framework for architectures provided by the CogAff schema. This approach can integrate a broad range of emotional processes including: the formation of a set of richer internal representations; and loss of control that can occur in emotional episodes.
Keywords :
"Computer architecture","Phylogeny","Biological system modeling","Genomics","Bioinformatics","Robots"
Conference_Titel :
Neural Networks (IJCNN), 2015 International Joint Conference on
Electronic_ISBN :
2161-4407
DOI :
10.1109/IJCNN.2015.7280431