Intelligent learning agent for collaborative virtual workspace

Abstract

PurposeIntelligent agents are becoming an essential part of collaborative virtual environments. The purpose of this paper is to present an architecture of a learning agent that is able to utilize machine learning techniques to monitor the user's actions.Design/methodology/approachA learning agent is developed and integrated into federated collaborative virtual workspace.FindingsThe experimental results showed that the combination of genetic algorithms and reinforcement learning algorithms provides the agent with better learning capability resulting in better predictions for the user.Originality/valueThis paper provides experimental results and a performance analysis in terms of accuracy of predictions, processing time, and memory utilization of the agent.

FCVW

“…The purpose of the learning agent is to monitor Notes used by the FCVW user who owns the agent, and to try to predict Notes in which the user might be interested in the future. To create such an agent, the user sets various parameters and attributes using the learning agent interface (Figure 21) and the agent then uses these parameters and re-enforcement learning to perform its task (Shakshuki, E. & Matin, R., 2008).…”