To the FIPA Home Page Publication Details
Up
Pub By Author
Pub By Year
Submission
Details

 

Name

E-Learning Agents

Date Published

2002-10-28

Reference

Presented during MALCEB 2002 (Erfurt Germany)

Keywords

Multi-Agent Systems, E-Learning, Knowledge Management, Skill Management, XML, Learning Object Metadata

Authors

Alfredo Garro and Luigi Palopoli

Abstract

E-learning is nowadays recognized as one of the key components of Enterprise Knowledge Managemnet platforms. Given a project specification, the platformshould be able to suggest a project team, to measure human rsources competence gaps and to comtribute to reduce them by creating personalized learning paths. In this paper we propose an XML based Multi-Agent System to perform the following tasks: (i) supporting Chief Learning Officers in defining roles, associated competencies and knowledge level required; (ii) mknaging skill map of the organization; (iii) measuring human resources competence gapx; (iv) supporting employees in filling their competence gaps as related to their roles; (v) enriching a given courseware or creating personalized learning paths according to feedbacks user provedes in order to optimize the acquisition of needed competencies; (vi) assisting Chief Learning Offficers in choosing the most appropriate employee for a given role.

URL

http://www.isi.cs.cnr.it/isi/garro/masel.pdf

Up

Name

F-One Racing-MAS & the Borda Count Interaction Protocol

Date Published

2002-11-21

Reference

Hopmans, G. , F-One Racing-MAS & the Borda Count Interaction Protocol, Co-operating agents to achieve a collective goal in Formula One Racing. Master of Science Thesis, University of Maastricht, November 2002

Keywords

Borda Count, Consensus mechanism, Interaction

Authors

Gabriel Hopmans

Abstract

The F-ONE RACING-MAS project examines how agents can co-operate to achieve an efficient strategy for a F1Racing team. These agents use complex interactions composed by many messages for their inter-working. FIPA is a foundation, formed with the goal of producing software standards to enable this kind of inter-working between inter-acting agents. One of the generic tools the FIPA has specified is the Interaction Protocol (IP). Interaction Protocols are used to design agents interaction providing a sequence of acceptable messages and a semantic for those messages [Poggi et al., 2000a]. However several studies and projects, including the recent, have shown examples wherein the FIPA standard IP's are not suitable for their particular scenario (for instance [Bergenti & Somacher, 2000]). Furthermore more FIPA-members are claiming that they should evolve more in IP's. [Levine, 2001] In the scenario for F-ONE RACING-MAS, the F1Racing team experts have to agree upon a common strategy over their own preferences. This thesis has examined the FIPA IP's and previous research on voting schemas and rules. Several papers have presented voting schemes, which allows agents to arrive at a consensus choice while balancing different user preferences ([Sen et al., 1997]) or whereby the system provides satisfactory recommendations ([Mukherjee et al., 2001]). The most promising protocol for F-ONE RACING-MAS is the Borda Count protocol. The Borda mechanism uses all the information provided by the voters and appears to be the unique voting method to represent the true wishes of the voters [Saari, 1994]. The thesis presents a proposal to extend the standard FIPA IP's with a new protocol that encompasses the functionality of the Borda Count mechanism. This contribution satisfies the minimal FIPA criteria for the proposed IP to be FIPA compliant (like an AUML-representation). Further research has to examine extensions for this Borda Count protoc!
ol, like an iteration-based approach for reaching compromises in agent systems.

URL

http://www.crs.unimaas.nl/g.hopmans/BordaCount/Hopmans2002.zip

 

 

Up

Name

IASTED International Conference SEA

Date Published

November 9 - 11, 2004

Reference

MIT, Cambridge, MA, USA

Keywords

Agent Oriented Software Engineering (AOSE), AUML, super role, atomic role, agent class, role class

Authors

Salaheddin J. Juneidi, George Vouros

Abstract

Software engineering development is crucial for industrial and commercial applications as these systems are required to operate in increasingly complex, distributed, open, dynamic, unpredictable, and inherently highly interactive environments.  This work is being motivated by the need to engineer complex systems with autonomous entities, to manage their inherent complexitiy during analysis, design and implementation, towards the development of systems with adaptations abilities. The paper introduces the Agent Role Locking (ARL) theory that provides a new conceptualization of the relation between agents, roles and objects. ARL calls on modification of AUML agent class diagram by introducing role classes in conjunction to agent classes and preserves the distinguishing characteristics between agent and object entities.

URL

http://www.icsd.aegean.gr/incosys/Publications/index.html

E-mail Contact

jneidi@aegean.gr

Up

Name

Lecture Notes in AI

Date Published

2002

Reference

[39] Buckle P, Moore T, Robertshaw S, Treadway A, Tarkoma S and Poslad S.  Scalability in Multi-Agent Systems: the FIPA-OS Perspective. LNAI Vol 2403, UKMAS proceedings, Springer Verlag 2002, 110-130, ISB 3-540-43962-5.

Keywords

scalability, fipa-os

Authors

Buckle P, Moore T, Robertshaw S. Treadway A, Tarkoma S and Poslad S.

Abstract

As agent systems move out of the research laboratories towards commercial application environments it is becoming increasingly apparent that scalability issues have to be investigated.  A revision of what scalability means is proposed; especially regarding its application in the description of attributes of agent systesm is described, and investigated through discussing qualitative as well as uantiative issues. A case study of a current project is provided to demonstrate how measures that determine how agent scalability might eventually be measured ispresented

URL

http://www2.elec.qmul.ac.uk/~stefan

Up

Name

MATA’02 4th Int. Worskhop on Mobile Agents for Telecommunications Applications, MA2002 6th IEEE Int. Conf. on Mobile Agents, Barcelona, 23-24 October 2002

Date Published

2002

Reference

(Keynote speech) Poslad S. Security, Trust & Privacy for Multi-agent, Multi-Domain (MAMD) applicaitons.  MATA’02 4th Int. Workshop on Mobile Agents for Telecommunications Applications, MA2002 6th IEEE Int. Conf. on Mobile Agents, Barcelona, 23-24 October 2002

Keywords

trust, security, privacy, multi-domains

Authors

Poslad S.

Abstract

We live in an information age where we have access to a huge variety of information and where we have a range of communication options to access and share this information. Software agents which exhibit behaviour such as autonomy, social ability, reactivity, pro-activeness and mobility are a powerful paradigm to promote more automatic, more manageable and semantically richer information access and sharing.
Security, privacy and trust are desirable and often necessary properties for these information systems. Although human agents could use the basic security features in the underlying communication and computer infrastructure without deliberation, human agents find it useful to reason about security. They can adapt their communication with respect to the environment and to the available resources, for example, to restrict eavesdropping or select communication channels that support greater degrees of privacy depending on the application and context. The same is true of software agents that are embedded in complex distributed information systems.
This presentation discusses the work being done by the FIPA (Foundation for Intelligent Physical Agents) forum in the area of multi-agent security. Multi-agent security is examined from two major perspectives: how to make agent based systems secure and how to use agents and their idiosyncrasies to enhance the security of agent and non-agent based systems.

URL

http://www2.elec.qmul.ac.uk/~stefan

Up

Name

Referencing Objects in FIPA SL:  An Analysis and Proposal

Date Published

15 July 2003

Reference

Proceedings of the 2nd International Workshop on Challenges in Open Agent Environments at AAMAS 2003

Keywords

Object-oriented ontologies. References, FIPA SL, Semantics

Authors

Stephen Cranefield and Martin Purvis.

Abstract

Although the syntax and semantics of mainstream agent content languages are based on those of predicate logic, the popularity of the Java programming language, the availability of various free Java-based agent development toolkits and the use of frame-based ontology modelling languages have meant that many developers of multi-agent systems are accustomed to conceptualising their problem domain in terms of classes and objects.
This paper examines the current practice of referring to objects within FIPA SL expressions using functional terms, with particular focus on how well this corresponds to the semantics of the language. After concluding that this use of functional terms is not correct, an extension to the syntax for identifying reference expressions is proposed. This notation allows objects to be referenced in terms of their attribute values in a well defined way, while requiring minimal change to current patterns of use.

URL

http://www.agentcities.org/Challenge03/Proc/Papers/ch03_cranefield.p df

E-mail Contact

scranefield@infoscience.otago.ac.nz

Up

 

© 2005 Foundation for Intelligent Physical Agents. All Rights Reserved.

Please send comments to webmaster(at)fipa.org