il giorno 29 novembre si terrà una miniscuola su tematiche relative ad oggetti ed agenti.
Le iscrizioni sono ormai chiuse.
Agent-Oriented Software Engineering
Federico Bergenti, Paola Turci
Introduction and concepts
What is AOSE?
Basic abstractions and metaphors of AOSE
AOSE vs. SE
What is a methodology?
A short introduction to GAIA
A short introduction to DESIRE
Tools for AOSE and conclusions
Modelling MASs with AUML
Realizing MASs with JADE
Current trends and future directions of AOSE
Programming multi-agent systems with JADE
Fabio Bellifemine*, Giovanni Caire
* non sarà presente alla scuola
JADE (Java Agent DEvelopment Framework) is a software framework fully implemented in Java. It allows reducing the time-to-market for developing distributed multi-agent applications by providing a set of ready and easy-to-use functionalities that comply with the standard FIPA specifications and a set of tools that supports the debugging and monitoring phases. JADE has been widely used over the last years by many organizations (both industrial and academic) in a variety of contexts: research projects, industrial prototyping, tutorials, and as a teaching support for agent-related courses in many Universities all over the world. JADE is extremely versatile and its features and footprint can also fit the constraints of environments with limited resources, such as java-enabled mobile phone devices J2ME-CLDC MIDP 1.0. It has been also integrated into complex architectures such as .NET or J2EE where JADE becomes a service to execute multi-party proactive applications.
JADE is a free software distributed by TILAB, the copyright holder, in open source under the terms of the LGPL (Lesser General Public License Version 2). Since May 2003, a JADE Board has been created that supervises the management of the project. Currently the JADE Board consists of 5 members: TILAB, Motorola, Whitestein Technologies AG, Profactor, France Telecom. The latest version of JADE is JADE 3.2, which has been released on July 2004. The aim of this tutorial is to provide a walk through all the main functionalities provided by JADE; a set of examples will show how to concretely use these functionalities to build agent-based applications.
Implementing the Proactive Behavior in Intelligent Agents: The Role of Abstraction
The introduction of agents has prompted investigations on AI algorithms and techniques, with the aim of using them in real-world domains, as classical assumptions no longer hold for them. In particular, real-world domains are usually not-completely controllable, not-completely accessible, and dynamic. An idealized agent devised for dealing with these issues should exhibit, other than social ability, proactive, reactive, and adaptive behavior. Moreover, according to the experimental evidence observed on the way humans conceptualize and solve problems, some abstraction abilities could be desirable. This talk addresses the problem of how to implement the proactive behavior in an intelligent agent, from both a theoretical and a pragmatic perspective. The talk starts with presenting general concepts of agent architectures, with particular emphasis on the implemented behaviors. The second part of the talk focuses on classical planning, resorting to toy problems for illustrating the underlying relevant issues. Furthermore, some abstraction techniques useful to significantly reduce the time complexity of the overall search are proposed. In the third part of the talk, planning techniques devoted to deal with the complexity of real-world domains are discussed. A case study is illustrated throughout the talk to evaluate the impact of no-classical assumptions on the complexity of the planning task. Finally, relevant issues related to the problem of how to coordinate and distribute planning tasks are briefly depicted.
Fabio L. Bellifemine is a senior project manager at the Department of Services and Multi-media of TILAB, Torino. He graduated in Computer Science from the University of Torino in 1988 and, prior to joining TILAB, until 1994 he held a researcher position at the Italian National Research Council. Since 1997, he is interested in the multi-agent system research and he is involved in the FIPA standardization body where he currently chairs the FIPA Architecture Board. In 1999 and 2002 he received a diploma from FIPA for its outstanding contribution to the activity. He is the leader of the JADE project and president of the JADE Board.
Federico Bergenti is a researcher at CNIT (Consorzio Nazionale Interuniversitario per le Telecomunicazioni) since 2001. He is a contract Professor at the University of Parma since 2001 for the course Ingegneria del Software A. His research interests focus on agent and object-oriented technologies and their use to develop reusable and maintainable software systems. He received a Ph.D. in Information Technology (2001) and a Laurea degree in Electronics Engineering (1998) both from the University of Parma.
Giovanni Caire received his Degree in Engineering at Politecnico di Torino in 1992. He joined CSELT in 1993 where he started working within the End-user Services division - Multimedia and Video Services department. He was engaged in several projects in the field of audio visual technologies and in particular he was the coordinator of the European project ATMAN that dealt with the trading of audio visual contents. Since 1998 his interest is in the field of intelligent agent technology. He has been actively participating in the design and implementation of the JADE agent development framework, in compliance with the FIPA Specifications. In 2001, within the scope of the European project LEAP, he leaded the group that successfully accomplished the porting of JADE on wireless devices. Currently he plays the role of Technical Leader in the JADE Governing Board, an international organization with the mission of promoting the evolution and the adoption of JADE by the mobile telecommunications industry.
Paola Turci received her Laurea degree in Electronics Engineering from the University of Bologna. She received her Ph.D. in Information Technology from the Department of Information Engineering of the University of Parma with a dissertation on Agent-Oriented Software Engineering. From 2003, she is contract Professor of the Faculty of Engineering of the University of Parma where she teaches Foundations of Computer Science and Web-oriented Systems. Her primary research activities are in the areas of multi-agent systems, object-oriented programming and software engineering.
Eloisa Vargiu obtained her MS degree and PhD in Electronic and Computer Engineering from the University of Cagliari, Italy, in 1999 and 2003, respectively. Since 2000, she collaborates with the Intelligent Agents and Soft-Computing (IASC) Group at the Dept. of Electrical and Electronic Engineering (DIEE), University of Cagliari (the IASC group is member of the European Network of Excellence in AI Planning, and of the European Network of Excellence for Agent-Based Computing). Her educational activities include tutoring in several Computer Science courses (Fundamentals of Computer Science, Software Engineering, Languages and Translators, Object Oriented Programming Languages). Since March 2003, she is an assistant researcher at the DIEE. Her research interests are in the field of intelligent agents; in particular, agent architectures, planning algorithms, and abstraction hierarchies. She is member of AAAI (American Association for Artificial Intelligence), AI*IA (Associazione Italiana per l'Intelligenza Artificiale), and MAS-AI*IA (Working Group of Agents and Multiagents Systems)