Romania

University "Politehnica" of Bucharest, AI-MAS laboratory

Membership Details (University/College)

Contact:
Member node: 117
Splaiul Independentei No. 313, sector 6, Bucuresti 060042, Romania
Romania

>>> Include yourself at this organisation
>>> Update your name or personal details listed below
>>> Submit papers, projects etc to the archive
>>> Update Member Description or Contact details

Description

University "Politehnica" of Bucharest is the biggest and oldest technical university in Romania. AI-MAS Group is part of the Computer Science Department at University "Politehnica" of Bucharest. Founded in 1997, our research focuses on multi-agent systems, with special interest in coordination mechanisms, automated negotiation, MAS architectures, and multi-agent learning. We are also concerned with conceptual models for on-line group learning, evolutionary agents, and models of affective computing.

The AI-MAS Group %2(http://turing.cs.pub.ro/ai_mas) has been member no. 021 in AgentLink II

Projects developed by AI-MAS Group

COOP is a negotiation system with self-interested agents that use evolved cooperation profiles to adjust negotiation strategies. The agents' behavior is motivated by gain and by the necessity to cooperate with other agents to achieve their goals. The negotiation strategy is represented explicitly, as a set of rules. The system is written in Java.

AGCOR is a system for organizational coordination using intelligent agents. The system allows the definition of several agents, ranging from totally cooperative to self-interested ones that achieve cooperation in order to share common gain. It is aimed at modeling virtual organizations. It is written in Java.

ARGUMENT is a system comprising a society of self-interested agents that use argumentation-based negotiation to reach agreements. The agents use different types of argumentation, for example qualitative arguments and arguments that are drawn from conventional negotiations based on costs and gain. It is written in Jade and Java.

AGENT-FISHBANK is a multi-agent system for solving the problem of rational exploitation of natural renewable resources by self-interested agents, known also as the Tragedy of Commons. The system is geared towards a particular instance of this problem, namely the FishBanks game. The agents in the system are hybrid ones, with genetic and cognitive components.

RLN is a framework for testing learning in negotiation. The agents are negotiating different negotiation situations and objects, including arguments in favor of successful contracts. The negotiator uses reinforcement learning to learn its negotiation strategy when faced with multiple negotiation primitives in repetitive encounters.

ICARUS is an intelligent interactive trading system for the stock market. The system comprises a collection of financial analysis tools and offers anticipative scenarios of stock evolution. Its basic architecture is a multi-agent system and prediction is based on reinforcement learning of stock trading evolution. The system is designed as an adaptive analysis tool for small and medium size brokerage companies.


Projects currently under development

MIRA (My IRrational Agent) aims at developing a model and an implementation of emotional agents. Emotions are viewed as a way of extending the classical BDIG model and have a definite role in the agent?s mechanisms of belief change, decision-making and planning. The project will be focused on both agents? emotion communication to the user and artificial modeling of user?s emotions. The project also aims at finding different approaches to identify user emotions, for example the user?s frustration.

ADEPT is a system targeted at developing an adaptive multi-agent system. The agents in the system will adapt their structure, behavior, autonomy, task distribution and resource allocation according to the problem to be solved and to the given environment. The system is conceived as an open one, in which agents have adjustable autonomy and in which communication and negotiation is adaptable, depending on new agents entering the system.

Course

[Course] Florea, Adina Magda (2005) Multi-agent Systems.

This list was generated on Mon Sep 10 16:39:27 BST 2007.

   

AgentLink is the European Commission's IST-funded Coordination Action for Agent-Based Computing
and is coordinated by the
University of Liverpool and University of Southampton
If you encounter any problems with these pages please contact web@agentlink.org.