Search

A virtual world is an online community in the form of a computer-based simulated environment, through which users can interact with one another and use and create objects. The non-player characters (NPC) in virtual world are following a fixed set of pre-programmed behaviors and lack the ability to adapt with the changing surrounding. Reinforcement learning agent is a way to deal with this problem. However, in a cooperative social environment, NPC should learn not only by trial and error, but also through cooperation by sharing information. The key investigation of this paper is: modeling the NPCs as multi-agent, and enable them to conduct cooperative learning, then speeding up the learning process. By using a fire fighting scenario in Robocup Rescue, our research shows that sharing information between cooperative agents will outperform independent agents who do not communicate during learning. The further work and some important issues of multi-agent reinforcement learning in virtual world will also be discussed in this paper. [ABSTRACT FROM AUTHOR]

In this paper we study the Airport Network of China (ANC), which represents China΄s domestic civil aviation infrastructure, as a complex network. We mine the structure and evolution of ANC over the past twenty years by using the real aviation data in the year of 1984, 1993 and 2006. The main contributions can be summarized as three-fold: firstly, we analyze ANC by using the complex network analysis method and find that ANC is a typical small world network with high clustering coefficient and small diameter; secondly, we find that the evolution of ANC over the past twenty years meets the densification law and shrinking/stabilizing diameter law; lastly, some interesting patterns of airports in ANC are found by the visual data mining, such as Circle Pattern, Province Capital Pattern and Star Pattern. [ABSTRACT FROM AUTHOR]

This paper presents a peer-to-peer based execution model to improve the availability and reliability of Geospatial Web Services (GWS) composition. A service community concept and a QoS model of GWS are proposed to discover the best quality GWS engines in the peer-to-peer network. The engines coordinate other participants in charge of initiating, controlling, monitoring the associated GWS execution, and adopt an alternative approach for failure recovery. The paper also presents the implementation of a peer-to-peer GWS composition system prototype based on JXTA platform. [ABSTRACT FROM AUTHOR]