António Luís Lopes

Abstract: The distributed nature of peer-to-peer networks offers a solid ground for the deployment of environments where multiple agents, managing several resources, can cooperate in pursuing common and individual goals while achieving good overall performance. In this article we present a survey of recent work on the integration of multi-agent systems and peer-to-peer computing for resource coordination (including discovery, composition and execution of resources) and we propose an approach for optimizing resource coordination through the use of efficient peer-to-peer search mechanisms relying upon a powerful semantic overlay network. We also present an approach for the dynamic development of the required semantic overlay network from a network of randomly-connected peers.

Abstract:

Abstract:

Abstract: The need to add semantic information to web-accessible services has created a growing research activity in this area. Standard initiatives such as OWL-S and WSDL enable the automation of discovery, composition and execution of semantic web services, i.e. they create a Semantic Web, such that computer programs or agents can implement an open, reliable, large-scale dynamic network of Web Services. This paper presents the research on agent technology development for context-aware execution of semantic web services, more specifically, the development of the Service Execution Agent (SEA). SEA uses context information to adapt the semantic web services execution process to a specific situation, thus improving its effectiveness and providing a faster and better service to its clients. Preliminary results show that context-awareness (e.g., the introduction of context information) in a service execution environment can speed up the execution process, in spite of the overhead that it is introduced by the agents’ communication and processing of context information.

Abstract: This paper proposes to use cognitive multi-agent based simulation to facilitate human team formation decision processes. The models of organizational actors are acquired by a data-gathering monitoring system that stores all the interactions that take place in the selected cooperative task. In this knowledge acquisition process, only a relatively small number of team compositions are monitored. The extracted models are used in the agent-based simulation of all the possible team compositions. Simulated team compositions are evaluated and only a small set of the most promising ones is submitted to the human decision maker. An emergence model overrides the multi-agent based simulation in localized situations for which the involved set of agents should not be trusted. The model of the simulation agents is a motivated case-based reasoning system, in which the regular functioning of the CBR is biased to fulfill the agent's motives. The motivation model includes individual motivation and social influence.

Abstract: This paper describes Chomsky, a content language translation agent. This agent provides a service, which first translates the content expression from its original language into an abstract logic language (ALL). Then, the resulting ALL expression is translated to the desired target language. ALL has been designed as a superset of most known content languages to avoid loosing expressiveness during translation. For more than three supported languages, using an intermediate language approach involves fewer translators than a pair wise approach. Currently, Chomsky supports FIPA-SL, KIF, and Prolog Content Language. In one mode of operation, Chomsky returns the result of the translation request to the client agent. In the other mode, Chomsky receives a message from the intended sender whose content is expressed in language L1, it translates the content to the content language used by the intended receiver (L2), and it sends the message with the L2 content to the intended receiver.

Abstract: The research project CASCOM (Context-aware Business Application Service Co-ordination in Mobile Computing Environments) will implement, validate, and trial value-added support for business services for mobile workers and users across mobile and xed networks. The vision of the CASCOM approach is that ubiquitous application services are flexibly co-ordinated and pervasively provided to the mobile users by intelligent agents in dynamically changing contexts of open, large-scale, pervasive environments.

Abstract: This paper describes an object-oriented framework for ontology representation, which was derived from the analysis of a set of use cases for ontology services provided by ontology agents in open agent networks. The described use cases are converted into a set of requirements for the Information System supporting the ontology service.Each ontology on the server has classes, properties and methods. Classes have attributes, which are properties, methods and axioms. Attributes have facets. Methods have arguments, return value and method definition through axioms. Classes and properties may be arranged in hierarchies. The proposed framework accommodates the expression of arbitrary axioms about the entities in the ontology. We propose a situation calculus approach for describing the action methods of the ontology. Action methods are described by state change and state constraint axioms. The frame problem is handled assuming that nothing changes unless explicitly stated.We present the definition of several relations between entities of different ontologies that allow inferring global relations between those ontologies. Finally, we propose Extended-SL as the Representation Language of O3F.

Abstract: This paper presents The Pagoda of Creation, a general web-based system for automatic creation of Personal Assistants for a worldwide agent network. Our vision is an agent network, as widespread as the WWW, and as easy to access as launching a browser in a laptop.Agent networks will be more intelligent than the current Internet but they risk being used only by specialized users. In fact, in order to access the Web, the user just needs to use a browser. But if s/he wants to interact with an agent network, s/he will need personal assistant agents (PA). Undoubtedly, writing a PA does require highly specialized software skills.The Pagoda of Creation is an intelligent web-based system runnable from any browser that easily creates personal assistant agents that will enable non-specialized users to have access to networked societies of agents operating in diverse domains. The created Personal Assistants may have web-based user interfaces (among other possibilities), personalized for the user and for the chosen application domain.Personal Assistants will be created from a database of agent templates for a diverse range of application types, accurately described using well-established modeling and representation languages. An agent template contains a model of the application domain, a formal interface specification, a description of the top-level control flow, and decision knowledge.Developers of new agent network application types will provide PA templates for their application types. These templates can be generated using interface tools also provided by the described Web-system.

Abstract: Collaborative environments, where multiple heterogeneous agents (managing several resources) can coop- erate in pursuing common and individual goals, are a step forward in creating real-world agent societies. However, current research in agent negotiation and in service coordination is still not enough for building such an agent-based society, capable of jointly solving complex planning problems and still achieve overall good performance. Most often, current work relies on either some centralised component or pre-defined social structure, which can compromise the system in terms of scalability, openness and robustness, and fails to address general problems. By using efficient network search algorithms and network evolution techniques it is possible to build and maintain a semantic overlay network from a totally unstructured distributed network, which in turn will simplify and optimize the distributed planning process amongst heterogeneous agents. We developed distributed versions of well-known planners that operate on top of the referred semantic overlay network and through a set of tests (using different scenarios) we were able to determine which is the best algorithm.

Abstract: A semantic overlay network is a powerful mechanism for collaborative environments where multiple agents, managing several resources, can cooperate in pursuing common and individual goals while achieving good overall performance. However, building such a social structure dynamically from an unstructured peer-to-peer network is a lengthy process if appropriate algorithms and techniques are not used. In this paper, we analyse a set of network evolution techniques that improve the performance of classic approaches, such as the flooding search algorithm. We compare the efficiency of these enhanced classic algorithms with our previously proposed search algorithm, which has also been improved through the referred techniques. Evaluation tests show that the improved version of our algorithm outperforms the improved version of the classic search algorithm and efficiently creates a semantic overlay network for agent-based resource coordination.

Abstract: This paper presents the research on agent technology development for context-aware execution of semantic web services, more specifically, the development of SEA (Service Execution Agent), a semantic web services execution agent that uses context information to adapt the execution process to a specific situation, thus improving its effectiveness and providing a faster and better service. Preliminary results show that the introduction of context information and context-aware capabilities in a service execution environment can speed up the execution process, in spite of the overhead that it is introduced by the agents’ communication and processing of context information. The developed service execution agent uses standards such as OWL-S service descriptions and WSDL grounding information. Also, an AgentGrounding definition has been proposed to enable the execution of semantic web services provided by agents.

Abstract: This paper describes the initial PhD research work for developing an innovative process for agent coordination in unstructured distributed environments. In this proposal, an agent that receives a request will use an AI-based planning algorithm to decompose the received request into several subtasks; it will handle those for which it has know-how and available resources; and it will use decentralized networks computing methods to delegate the rest of the request to other agents that can possibly handle it. These, in turn, will recursively apply the same process until the initial request is completely solved or until some specified termination condition holds. The work contributions to the state-of-the-art are: the creation of an AI-based planning algorithm capable of partial task decomposition with only partial knowledge of available planning operators; and the creation of a coordination infrastructure that enables agents to delegate subtasks to other agents in totally unstructured decentralized environments.

Abstract: In this paper, we present a general architecture for service delivery and coordination in intelligent peer-to-peer (IP2P) environments that has been developed within the CASCOM research project. Our essential approach is an innovative combination of agent technology, Semantic Web Services, peer-to-peer, context-awareness, and mobile computing for intelligent peer-to-peer mobile service environments. Services are provided by software agents exploiting the coordination infrastructure to efficiently operate in highly dynamic environments. Our infrastructure includes efficient communication means, support for context-aware adaptation techniques, as well as dynamic service discovery and composition planning. For end users, the architecture provides seamless access to Semantic Web Services anytime, anywhere, and using any device. Our architecture is being evaluated using a sample ad-hoc emergency healthcare assistance application scenario. We deployed a prototype of an open IP2P service environment and expect results on methods for service provision, discovery, composition, and monitoring in mobile environments.

Abstract: Emergency Health-Care based scenarios provide the motivation to develop supporting technologies for dealing with situations where people need medical assistance because of a sudden disease or emergency. The innovative combination of intelligent agent technology, Semantic Web services, peer-to-peer, and mobile computing for intelligent peer-to-peer mobile service environments is the corner stone of the CASCOM project, which aims at providing a value-added support for business services for mobile workers and users across mobile and fixed networks, especially for e-health environments. In this paper, we describe the technical approach developed in the CASCOM project for the context-aware secure composition planning and execution of semantic web services.

Abstract: Semantic Web Services are the driving technology of today’s Internet as they can provide valuable information on-the-fly to users everywhere. Standard initiatives such as OWL-S and WSDL enable the automation of discovery, composition and execution of semantic web services, i.e. they create a Semantic Web, such that computer programs or agents can implement a reliable, large-scale interoperation of Web Services. This paper describes the on-going research on agent technology development, more specifically, the creation of an intelligent agent for context-aware execution of semantic web services, allowing the adaptation of the execution process to a specific situation, thus improving its effectiveness. We introduce the concept of context-awareness in semantic web services’ execution to better represent the dynamics of P2P networks, which is the environment on which the service execution agent will operate. The context-awareness feature is provided by a generic con- text framework. We have decided to adopt the agent paradigm in this problem because we intend to integrate this work in open societies of agents, enabling these to discover, compose and execute semantic web services. The same approach has been used for service discovery and mediation of semantic web services. The agent will be implemented using component-based software and by extending (overcoming current limitations on) agent technology, such as JADE and semantic web services execution mechanisms, such as OWL-S API.

Abstract: The research project CASCOM will implement,validate,and trial value-added support for business services for mobile workers and users across mobile and fixed networks. The vision of the CASCOM approach is that ubiquitous application services are flexibly co-ordinated and pervasively provided to the mobile users by intelligent agents in dynamically changing contexts of open,large-scale, pervasive environments. The essential approach of CASCOM is the innovative combination of intelligent agent technology, semantic Web services, peer-to-peer, and mobile computing for intelligent peer-to-peer (IP2P) service environments. The services are provided by software agents exploiting the co-ordination infrastructure to efficiently operate in highly dynamic environments.

Abstract: This paper presents the on-going research on agent technology development for context-aware execution of semantic web services, more specifically, the development of SEA (Service Execution Agent), a semantic web services execution agent that uses context information to adapt the execution process to a specific situation, thus improving its effectiveness and providing a better service. The agent uses standards such as OWL-S service descriptions and WSDL grounding information. Also, an Agent Grounding definition has been proposed to enable the execution of semantic web services provided by agents.

Abstract: This paper describes BAT (Brainstorm Assistance Tool), a computational system for supporting distributed brainstorms based on a well-defined structured graphical interface. BAT was created within the KnowledgeAndCulture.org project, whose objective is to create a set of tools to facilitate the human resources’ activity of team formation for cooperative tasks. In this case, the brainstorm-like electronic meeting was the chosen cooperative task. This paper describes BAT, which monitors and records the interactions that take place between several participants of distributed brainstorms. The data generated by BAT is used in agent-based simulation to assist the selection of the best team composition to perform specific cooperative tasks. The system can be used through a graphical language defined for general purpose Brainstorms or through the addition of plug-ins for domain-specific Brainstorms. In this paper we present the overall system design, including the underlying brainstorm model and the output generation process.

Abstract: This paper analyses the specification of agent control and agent knowledge in DAML-S and its conversion into executable code. Instead of the usual XML syntax, we propose a S-Expression syntax of DAML-S, which facilitates introducing two extensions in DAML-S specification: concrete definitions of concepts used in DAML-S service descriptions, which were not defined in the DAML-S specification; and logic-programming constructs that may be used in service description. Two approaches are discussed with respect to the generation of executable code. Both of them contain a first step in which the DAML-S service-description is parsed into an appropriate computer program data structure, called the DAML-S description parse tree. The first approach converts the DAML-S description parse tree into source files that must be compiled and linked in order to create the executable agent. The second approach relies on the run-time interpretation of the DAML-S description parse tree.