Towards Basic Motivations for Autonomous Agents

摘要

This paper discusses our research towards development of a domain-independent autonomous agent that can learn and adapt to function in a domain. For an agent to work autonomously in a domain, it needs some motivation to get familiar with its environment. Motivation depends on a set of emotions and their state at any given time instance. In this paper we propose an agent model with two emotions that can motivate the agent effectively to perform a function. The model was developed and tested in a virtual world and their results are discussed. This paper concludes that our algorithm can perform its basic goals successfully without explicit domain based knowledge.