[an error occurred while processing this directive]
Only when sage warriors use intelligent agents
will they certainly have success.
-- Sun Tzu, The Art of War


Volume 1,Number 17
December 16, 1996
Baltimore, MD

Modified Monday, 16-Dec-1996 10:34:03 EST


Internet Publish and Subscribe Protocol

TIBCO Inc., and more than a dozen Internet companies have endorsed a proposed new industry standard for the "push" model of information distribution over the Internet. The proposed standard, called publish and subscribe, will reduce Internet traffic and make it easier to find and receive information on-line. The companies, which include Cisco Systems, Inc., CyberCash, Informix, Infoseek, JavaSoft, Sun Microsystems, Verisign, NETCOM, and others in addition to TIBCO, announced plans, products and support for publish and subscribe. TIBCO and Cisco Systems are developing an open reference specification for publish and subscribe technology. MORE. Tibco also announced Tibnet -- a family of products for implementing publish-and-subscribe in hardware and software and based on their Information BusŪ (TIBŪ) middleware software (e.g., Rendezvous). This supports both real-time messaging and message queueing for guaranteed delivery and it offers a tightly integrated CORBA 2.0 compliant Object Request Broker.

Software hunts and kills Net viruses

The Sunday Times (1 Dec 1996). VIRUS-KILLING computer software that uses artificial intelligence to find and destroy new viruses is to be set loose on the Internet by IBM next week. The software, originally developed to play backgammon, will spread itself through the Net over the next year, learning how to kill new strains of virus as it goes. According to Gregory Sorkin, a researcher at IBM's Watson research laboratory, the system will be far more successful than humans at fighting computer viruses. "Once it learns the viruses already out there, the system will even be able to predict what new viruses will appear, and work out ways of stopping them before they even exist," says Sorkin. The system uses temporal difference, a method which relies on the computer looking for patterns within virus software, rather than individual lines of program code.



Vince Delle Donne, Daniel Gauvin and Herve Marchal of CRIM (Computer Research center In Montreal) have developed LALO, a language and a framework for developing intelligent multiagents systems. LALO is a language which uses the Agent Oriented Programming (AOP) paradigm as defined by Yoav Shoam. In AOP, an agent is determined by it beliefs, his capabilities, and his commitments, which together comprise its mental state. AOP encourages a social view of computation in which communities of agents interact by exchanging information, sending specific requests, offering services, accepting or refusing tasks, competing with each other for a task to be accomplished or cooperating with each other. A program written in LALO is translated into C++ source code, and then may be compiled with your regular C++ compiler. The agents communicate with KQML (Knowledge Query Manipulation Language). The LALO framework is available on UNIX platforms, Windows NT and Windows 95. Send email to lalo@crim.ca for more information.

ILOG announces engine for C++ intelligent agents

ILOG announces the world's first dynamically adaptable engine for C++ intelligent agents. Paris -- October 30, 1996. ILOG, the world's leading provider of C++ class libraries for strategic business applications, today announced version 4.0 of ILOG Rules, its system for creating intelligent agents using high-performance telecoms, financial and help-desk applications. The new version comprises a unique feature that allows users to alter agent behavior as their needs change, without having to write C++ code. They can simply change existing rules or add new rules while the application is running. The Rules 4.0 interpreter allows an intelligent agent to load new rules directly from text, making it easier to implement real-time filters and batch checking systems. Complex systems involving multiple ILOG Rules intelligent agents, databases and feedback loops are now practical to implement. MORE .


Firefly meets Yahoo

Firefly Network Inc. and Yahoo have launched a public beta of a website recommendation service using Automated Collaborative Filtering (ACF) technology. It allows users to find websites they are likely to find of interest and to find other people who share similar interests. To access the service, go to My Yahoo and then click [Firefly/My Agent] button at the top right of the page. They say
"Out of a constantly expanding universe of information, firefly will recommend sites you'll like and connect you to people who like what you like. First you give your opinion of a range of Weqb sites, then firefly finds other people who share those interests. You can interact with these neighbors, or have firefly recommend sites based on what they and you like. It's an automated, worldwide word-of-mouth system."
Automated Collaborative Filtering is a technique for providing recommendations (or dis-recommendations) based on statistical matches of peoples' evaluations of a set of objects in some domain. It has been developed and explored at the MIT Media Lab and used in a number of recommendation systems, such as HOMR, WEBHOUND and Firefly. Typical domains might be movies, books, web pages, or TV shows. An evaluation of an object is typically a simple scalar value, such as an integer from zero to ten or a real between 0.0 and 1.0, and might represent the degree to which the person "likes" the object. An ACF system maintains a database of users and keeps records of items they have evaluated. The set evaluations a person has made form a vector and represents that person's opinions. These evaluation vectors are compared to the vectors provided by other users and people with similar opinions can be discovered. From this, one can predict the evaluations a person might make for new objects in the domain.


MORSE is a movie recommendation system done at BT Labs, Martlesham Heath, and is funded by BT. It was done as a MSc project by Donald Fisk (currently at University of London) under the supervision of by Dr. Andy Downton of Essex University and Brian Tester of BT.


NSF program in Learning and Intelligent Systems

NSF has announced an opportunity for interdisciplinary research in Learning and Intelligent Systems (LIS). Six NSF Directorates will coordinate and manage the initiative through a special Committee with an appointed Coordinator. The LIS initiative seeks to stimulate interdisciplinary research that will unify experimentally and theoretically derived concepts related to learning and intelligent systems, and that will promote the use and development of information technologies in learning across a wide variety of fields. The long-range goal ia broad and has the potential to make significant contributions toward innovative applications. The initiative focuses on fundamental scientific and technological research undertaken in the rigorous and disciplined manner characteristic of NSF supported endeavors. The initiative ultimately should have a major impact on enhancing and supporting human intellectual and creative potential. Consequently, development of new scientific knowledge on learning and intelligent systems and its creative application to education and to learning technologies are an integral part of this solicitation.



The advance program for the First International Conference on Autonomous Agents (Agents '97) is available. The conference will be held at the Marriott Hotel, Marina del Rey on February 5-8, 1997.

Enterprise Distributed Object Computing

The First International Workshop on Enterprise Distributed Object Computing, Gold Coast, Australia, October 24-26, 1997. The topics of the Workshop include: Enterprise architectures and frameworks, ODP enterprise language and business modeling methodologies, Workflow for intra and inter-organizational applications, Trader and Intelligent Brokers, Distributed intelligent agents for enterprise systems, and The use of mobile computing in enterprise systems. Papers are due June 1, 1997.

Intelligent Animated Interface Agents

There will be a workshop on Animated Interface Agents: Making them Intelligent Nagoya, Japan, August 23-25, 1997 in conjunction with IJCAI-97. Workshop attendance will be limited to maximally 40 people, persons without a submission should contact the organizers as soon as possible. Submit extended abstracts, project/demo descriptions, videos, etc by 3rd March 1997.


GROUP `97 -- International Conference on Supporting Group Work, "The Integration Challenge". November 16-19, 1997, Phoenix, Arizona USA. Papers are due May 15, 1997. GROUP `97 is interested in topics related to computer-based systems which have an impact on groups, organizations and social networks. A strong emphasis of GROUP `97 is the integration of multiple computer-based tools and technologies. Relevant issues include design, implementation, deployment, evaluation, methodologies, and effect of these systems. ... we observe a growing influence of the Internet, mobile computing, agent systems, ubiquitous computing, and virtual reality. We can expect that these new technologies will also exert a large influence on group/organizational structures and processes. These new technologies are exciting in their own right, but their technological and organizational integration to support groups raises many interesting questions and is a challenging new research agenda....Technical Track:... * Computer supported collaboration and negotiation, * Coordination and workflow technology, * Cooperative knowledge management and organizational or group memory, * New technologies for the support of groups, e.g., agent technology, ubiquitous computing, virtual reality * Social networks and virtual communities, ...


Workflow and agents

Several papers relating to the use of agents for workflow were presented at CSCW and the Web, An international Workshop organized by ERCIM/W4G, Sankt Augustin, Germany, February 7-9, 1996.

Agents based Workflow: TRP Support Environment ( TSE ), Jin W. Chang and Colin T. Scott, Andersen Consulting. "... The core concept of the above TSE is based on the WWW and a distributed set of autonomous cooperative processes known as agents. These agents facilitate various parts of workflows in order to improve efficiency among participants and that of overall process. Although the original concept of an agent, "acting on behalf of someone else" is simply lost in current agent researches [1] in A.I., it is precisely this notion which TSE adopts. In other words, the agents in TSE largely act on behalf of participants, organizing their agenda, looking for important events ( such as a deliverable is complete ), and even advise necessary steps to accomplish in a given task. Although the most of workflow functionalities may be replaced with a more extensive agent server, the approach in TSE has adopted workflow for the following reasons. One is the readily available set of existing workflow tools and a coming standard from Workflow Management Coalition. The other is by utilizing an existing workflow tool, an agent server can simply be a facilitator agent providing interfaces between agents and workflow tool. ..."

WebFlow: Coordination and Workflow on the World Wide Web, A. Grasso and D. Pagani, Rank Xerox. "... WebFlow is based on the Coordination Language Facility (CLF) developed at RXRC Grenoble (Andreoli et al.). The CLF is a process-oriented middleware that offers primitives to coordinate object-oriented software components. The CLF allows to specify coordination patterns and processes using a declarative, rules-based scripting language. A CLF coordinator enacts a process by managing the consistent negotiation among heterogeneous, distributed service objects. The coordinator engages conversations with the objects according to a specialized protocol implemented on top of multiple transport layers: Corba and HTTP...."

Knowledge Acquisition for Knowledge-Based Systems

The Tenth Knowledge Acquisition for Knowledge-Based Systems Workshop was held in Banff, Canada, on November 9-14, 1996. Some papers relating to agents presented at the conference were: * KAoS: An Open Agent Architecture Supporting Reuse, Interoperabiliby, and Extensibility, Jeffrey M. Bradshaw; * Beliefs, Intentions and DESIRE, Frances Brazier, Barbara Dunin-Keplicz, Jan Treur, Rineke Verbrugge; * Compositional Modelling of Reflective Agents, Frances Brazier, Jan Treur; * The Emergence of Knowledge through Modeling and Management Processes in Societies of Adaptive Agents, Brian R. Gaines; * A Methodological Proposal for Multiagent Systems Development extending CommonKADS, Carlos A. Iglesias, Mercedes Garijo, José C. Gonzelez and Juan R. Velasco; * Software Agents Based on Formal KL-Models, Josefina Sierra Santibeez; * Using Ontologies in Multi-Agent Systems, Sabina Falasconi, Giordano Lanzola and Mario Stefanelli; * Understanding, Building, and Using Ontologies, Nicola Guarino; * Ontologies as Vehicles for Reuse: a mini-experiment, I. Laresgoiti, A. Anjewierden, A. Bernaras, J. Corera, A.Th. Schreiber and B.J. Wielinga; * ONIONS Methodology and the Ontological Commitment of Medical Ontology ON8.5, Geri Steve and Aldo Gangemi; * Toward Distributed Use of Large-Scale Ontologies, Bill Swartout, Ramesh Patil, Kevin Knight and Tom Russ; * Towards Principled Core Ontologies, Andre Valente and Joost Breuker; * The Ontolingua Server: a Tool for Collaborative Ontology Construction, Adam Farquhar, Richard Fikes and James Rice; * A Networked, Open Architecture Knowledge Management System, Brian R. Gaines and Mildred L. G. Shaw; * Reuse for knowledge-based systems and CORBA components, John H. Gennari, Adam R. Stein and Mark A. Musen.

The Design of Intelligent Agents: a Layered Approach

The Design of Intelligent Agents: a Layered Approach, Joerg P. Mueller, Springer-Verlag, Lecture Notes in Artificial Intelligence Volume 1177, December 1996. ISBN 3-540-62003-6. 227pp. Includes index. Price DM 48,00 softback. Intelligent agents are computer systems that can make rational decisions and act autonomously in real-time, open, multi-agent environments; agent technology has the potential to play a key role in the design and operation of complex distributed systems in the years to come. This monograph presents a comprehensive state-of-the-art survey of approaches to the design of intelligent agents. On the theoretical side, the author compiles a set of general requirements for autonomous interacting agents and provides an essential step towards understanding the principles of intelligent agents. On the practical side, the agent architecture INTERRAP is introduced; the detailed description and evaluation of the architecture is an ideal guideline and case study for software engineers and researchers who are faced with the task of building a real agent system. This book bridges the gap between theoretical foundations and practical implementations of intelligent agents. It addresses both active and novice researchers and is essential reading for anyone interested in applying this exciting new software technology.

Jasper: Communicating Information Agents for WWW

Jasper: Communicating Information Agents for WWW, John Davies, Richard Weeks, Mike Revett, BT Laboratories, January 1996. This paper discusses a distributed system of intelligent agents for performing information tasks over the Internet WorldWideWeb (W3) on behalf of a user or community of users. We describe how agents are used to store, retrieve, summarise and inform other agents about information found on W3. Most current W3 clients (Mosaic, Netscape, and so on) provide some means of storing pages of interest to the user. Typically, this is done by allowing the user to create a (possibly hierarchical) menu of names associated with particular URLs. While useful, this quickly becomes unwieldy when a reasonably large number of W3 pages are involved. The solution we adopt to this problem is to allow the user to access information by a much richer set of meta-information than simply names assigned to particular URLs. Given the vast amount of information available on W3, it is preferable to avoid the copying of information from its original location to a local server. The local storage of only relevant meta-information also addresses this issue. The meta-information used includes automatically extracted keywords and summary, as well as the document title, universal resource locator (URL) and date and time of access. This meta-information is then used to index on the actual information when a retrieval request is made. When an agent stores a page, it will also automatically inform other users who it considers will be interested of the page's discovery via an email message. Thus Jasper is a small step towards automating the original vision for W3 as a network which supports co-operative working and the sharing of information. In addition, Jasper agents will modify a user's profile as the type of information the user is accessing changes. Pages stored in Jasper can be posted to interest groups, creating shared information resources about specific topics.

From Animals to Animats

From Animals to Animats 4 - Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior, September 9th-13th, 1996, Cape Cod, Massachusetts, edited by Pattie Maes, Maja J. Mataric, Jean-Arcady Meyer, Jordan Pollack, and Stewart W. Wilson. From Animals to Animats 4 brings together the latest research at the frontier of an exciting new approach to understanding intelligence. The contributors represent a broad range of interests from artificial intelligence and robotics to ethology and the neurosciences. Unifying these approaches is the notion of "animat" - an artificial animal, either simulated by a computer or embodied in a robot, which must survive and adapt in progressively more challenging environments. The 66 contributions focus particularly on well-defined models, computer simulations, and built robots in order to help characterize and compare various principles and architectures capable of inducing adaptive behavior in real or artificial animals. Topics, all from the perspective of adaptive behavior, include: The Animat Approach to Adaptive Behavior, Perception and Motor Control, Action Selection and Behavioral Sequences, Internal World Models and Navigation, Motivation and Emotions, Learning, Evolution, Coevolution, and Collective Behavior. Complex Adaptive Systems series, A Bradford Book, October 1996, 600 pp., $65.00 paperback, ISBN 0-262-63178-4-3. MIT Press.
[an error occurred while processing this directive]