UMBC Research on Software Agents and Multi-Agent Systems

11/18/98


Click here to start



6 up PDF version


Table of Contents

UMBC Research on Software Agents and Multi-Agent Systems

Some Assumptions and Biases

The Agent Paradigm

Why is communication important?

Agent Communication

Some ACLs

The intentional level, BDI theories, speech acts and ACLs: How do they all fit together?

Agents and agencies

Some Agent Research at UMBC

UMBC agent research

CIIMPLEX EECOMS

Manufacturing Enterprise Integration

CIIMPLEX Architecture

CIIMPLEX MAS Architecture

Supply Chain Enterprise Application Integration Global Visibility and Control of Business Processes across Apps

Negotiation among agents in the Supply Chain

General Problems

Specific Objectives

Specific Approach

1 Negotiation primitives

Examples of negotiation primitives

Defining negotiation protocols

2 An XML-based content

Laptop description seen “by eye”

HTML laptop description

XML Laptop Description

3 Controlling negotiating agents

Adjustable Autonomy

Recent Results

Jackal A Communications Infrastructure for Java-based Multi-agent Systems

Jackal Architecture: Overview

Conversation-based Protocols

Current Conversation Specification

Using Colored Peter Net Specifications

Jackal Message Distributor

Typical Components of a Multi-agent System

KQML Naming Scheme (KNS)

CARROT: Cooperating Agent-based Routing and Retrieval of Text

Conclusions

Some key ideas

Prospects

For More Information

PPT Slide

CARROT: Cooperating Agent-based Routing and Retrieval of Text

PPT Slide

Broker and back-end agents

Telltale

n-grams vs. words

Desiderata for Corpus Metadata

Telltale user interface

VR based visualization of retrieval

VR Approach

Visualizing a document space

PPT Slide

Linneaus

Author: tim finin

Email: finin@umbc.edu

Home Page: http://umbc.edu/~finin