Catálogo de publicaciones - libros

The use of a complex form of language for the purpose of communication with others, to exchange ideas and thoughts, to express statements, wishes, goals, and plans, and to issue questions, commands and instructions, is one of the most important and distinguishing aspects humankind. If artificial agents want to participate in a co-operative way in human-agent interaction, they must be able – to a certain degree – to understand and interpret natural language, and translate commands and questions of a human user and map them onto a suitable set of their own primitive actions. In this paper, we outline a general framework and architecture for the development of natural language interfaces for artificial agents. We focus in this paper on task-related communication, in a scenario where the artificial agent performs actions in a co-operative work setting with human partners, who serve as instructors or supervisors. The main aim of this research is to provide a consistent and coherent formal framework for the representation of actions, which can be used for planning, reasoning, and action execution by the artificial agent, and at the same time can serve as a basis for analyzing and generating verbal expressions, i.e. commands, instructions and queries, issued by the human user. The suggested framework is derived from formal methods in knowledge representation, in particular description logics, and involves semantic and ontological descriptive elements taken from linguistics, computer science, and philosophy. Several prototypical agent systems have been developed based on this framework, including simulated agents like an interior design system and a household robot, and a speech controlled toy car as example of a physical agent.

In order to enable the intelligent agents to be aware of the dynamic business requirements and strategies, and cooperate with other agents in a stable and explicit way, a policy and contract extended BDI logic (called BGI logic) has been proposed, by integrating contracts and policies into traditional BDI model. On the basis of BGI logic, this paper proposes a model of agent architecture as a concrete realization of it, called PDC-agent. PDC-agent is an extension of traditional BDI agent, by adding a policy engine, a contract engine and a goal maintenance component into agent’s interpreter. Besides, PDC-agent has two characteristics. First, the operation of PDC-agent is based on an event-driven mechanism. Various events drive the components of the interpreter. Second, the representation of PDC-agent is on the basis of ontology.

Despite a great deal of research, there still exists a number of challenges before making agent-based computing a widely accepted paradigm in software engineering practice. In order to realize an engineering change in agent oriented software engineering, it’s necessary to turn agent oriented software abstractions into practical tools for facing the complexity of modern application areas. The paper presents a customizable development architecture for multi-agent systems, which can empower the developer to assemble a methodology tailored to the given project by putting appropriate meta models together, much like developers building applications from third party off-the-shelf components. To exemplify its feasibility and effectivity, the construction of C4I system is presented as a case study.

Gaia is one of the first agent-oriented software engineering methodologies, which explicitly takes social concepts into account. Yet Gaia would neither suffice to adequately develop adaptive system in open environments nor describe role relation and hierarchy of organizational rule. On the other hand, a variety of models employed by different methodologies limit the progress of MAS, so it is necessary to develop a unified framework which includes the models adopted by different methodologies. FRAG is proposed as a systematic methodology, which is the extension of role models of Gaia on the relation among function, role, and agent, for developing adaptive systems in open environments. Having compared several methodologies, the ORRA process is proposed as a universal process in role-based methodology starting from the purpose of different models adopted in MAS methodology. The Conference Management Case study is introduced to exemplify ORRA’s process and to show the use and effectiveness of FRAG’s models in the development of MAS.

With the rapid information growth on the Internet, web information collection is becoming increasingly important in many web applications, especially in search engines. The performance of web information collectors has a great influence on the quality of search engines, so when it comes to web spiders, we usually focus on their speed and accuracy. In this paper, we point out that customizability is also an important feature of a well-designed spider, which means spiders should be able to provide multi-modal services to satisfy different users with different requirements and preferences. And we have developed a parallel web spider system based on multi-agent techniques. It runs with high speed and high accuracy, and what’s the most important, it can provide its services in multiple perspectives and has good extensibility and personalized customizability.

To design intelligent agents for multi-agent applications, like auctions and negotiations, we need to first analyze how agents should interact in these applications. Game theory is a tool, which can be used. In game theory, decision-making often depends on probability and expected utility. However, decision makers usually violate the expected utility theory when there is risk in the choices. Instead, decision makers make decisions according to their attitudes towards risk. Also, reputations of other agents in making certain actions also affect decision-making. In this paper, we make use of risk attitude, reputation and utility for making decisions. We define the concepts of , , and a formalized way to find the risk strategy equilibrium in infinitely repeated games. Simulations show that players get higher payoff by using risk strategies than using other game theoretic strategies.

Resources allocation and tasks scheduling is key technology in grid computing system. The market-based resources allocation model is considered as a good one. In this paper, a resources allocation model, based on reverse auction, was proposed, and its mechanism and related pricing algorithms were designed. In this model, a resources consumer invites a public bidding on the basis of his deadline and budget, then a resources provider bids according to his load, and the bidder who bids the cheapest price will win the auction. Numerous simulating experiments based on our proposed model was conducted, the experiments showed that our model can satisfy a user’s QoS demand on deadline and budget, and have better performance in user utility, society utility, load-balance, job-completed rate than a commodity market-based resources allocation model.

This paper presents data grid system based on agent for an efficient interoperability of distributed data. Purpose of the data grid system is an efficient interoperability through sharing and exchanging data among distributed databases. Interoperability of distributed data is a basic condition for cooperation of existing legacy systems. More nodes for cooperation, better for distributed environment. On the other hand, we have to overcome heterogeneous characteristics of data and physical problems of network. Therefore, we use agent system which is independent, autonomous and mobile in distributed platform for solving this problem. As mobile Agent interacts asynchronously, it could lessen network traffic and support workload and distributed processing of services so that it is suitable for data grid system. Also we use XMDR(eXtended Metadata Registry) in ISO/IEC 11179 for sharing and exchanging data and overcome the heterogeneity of data. Therefore we improve availability and transparency of distributed data based on data grid middle- ware using proposed mobile agent and could construct data grid system with low cost.

Since mobile computation involves variants information and different behaviors, it is very complicated in mobile computation systems. In order to analysis mobile computation and understand the essential of mobile computing, in the paper we develop a denotational semantics for mobile computation. Here we take CLAIM, a computational language for autonomous, intelligent and mobile agents, as our object language because it characterizes the essential ingredients in mobile computation. After abstracting the syntax construction of CLAIM, we give its semantics description. This is achieved by structuring the semantics in layers working at three different levels: internal ambient, ambients and programs based on the basic concept: ambient. For each of these three levels, their semantics are defined in detail as well as the relationship between levels. Through our approach, we can also obtain an explicit model of behaviors in mobile computation.

This paper analyzes the distinctive characteristics of grid environments and proposes a novel immunity and mobile agent based intrusion detection for grid () model. Then, the concepts and formal definitions of , , , , and in the grid security domain are given. Besides, the mathematical models of , (mature monitoring agent), (memory monitoring agent) , (communicator agent), and (beating off agent) are established. The effects of several import parameters on system performance and detection efficiency in the model of dynamic memory MoA survival are analyzed and shown in the experiments. Our theoretical analysis and experimental results show the model which enhances detection efficiency and assures steady performance in immune-based IDS is a good solution to grid intrusion detection.