دانلود رایگان مقاله انگلیسی بازیگر مجازی با هوش هیجانی و اجتماعی - الزویر 2018

Abstract

This work continues the effort to design and test the cognitive architecture eBICA: a general model of emotionally biased behavior control and decision making, with the focus on social emotional relationships. The key building blocks of the model include dynamics of mutual appraisals of actors, bi-directionally linked to behavior, on the one hand, and on the other hand – to M-schemas, that establish normal behavior of two or more actors involved in a mutual relationship, such as partnership. To test the model, we implement a Virtual Actor embodied as an avatar in a specifically designed virtual environment, and use several paradigms of its social interaction with humans. We show that virtual environments and associated paradigms can be divided into a hierarchy, on top of which are paradigms with dynamically changing social relationships and roles. Using paradigms of this kind, we show that a virtual actor can be indistinguishable from a human participant in three dimensions: efficiency, believability and social acceptability. Some of the social measures demonstrate a higherthan-human performance, inspiring an idea of a new challenge for BICA and AI. The general model of a Virtual Actor presented here is anticipated to sparkle ideas of broad impacts on many practical areas.

4 Conclusions and Discussion

In conclusion, in this work we presented a new platform that can be used for social psychological experiments in virtual environments, with anonymous participation of humans and Virtual Actors. We also presented the study of an implemented Virtual Actor based on the eBICA model [13]. This model uses a semantic map of elements of behavior, together with M-schemas, to generate decision making and behavior of a Virtual Actor. Results of the study allow us to validate the model. Two experimental paradigms were implemented and used for empirical testing of the model. The outcomes are the following. 1. Virtual Actor is indistinguishable from a typical human participant based on its efficiency. 2. Virtual Actor is indistinguishable from a typical human participant based on its believability. 3. Virtual Actor is slightly worse than a typical human participant in social acceptability, based on Criterion 1. 4. Virtual Actor is better than a typical human in social acceptability, based Criterion 2. The overall conclusion is that the implemented Virtual Actor performs at a human level. This outcome constitutes a strong support for the designed model of a Virtual Actor, based on the further developed cognitive architecture eBICA. The main volume of data collected in this study automatically in the form of logs of all events happening in the environment is not presented here, due to the limited volume of this publication. Its analysis will provide us with new insights into the model, as well as being able to serve as an alternative, behavior-based, objective basis for evaluations of the three characteristics presented here. For example, social acceptability can be measured by the frequency of obtaining help from others. Analysis of these details will be presented elsewhere.