A free email alerter of the latest news items and associated URLs.
Subscribe FREE to El.pub Analytic
El.pub Analytic emailed as soon as it's published.
Agents are software artifacts that act on behalf of users. It is believed that they will be a major factor in overcoming information overload in the future by carrying out tasks in cyberspace on behalf of users. One of the best articles defining agents and offering a taxonomy is "Is is an Agent or just a Program?" by Atan Franklin and Art Graesser.
- AgentWeb, a site with a great deal of information about agents is run by the UMBC Laboratory for Advanced Information Technology. There are extensive pointers to other sites, online papers and other resources. They publish an electronic newsletter, AgentNews.
URL: AgentWeb http://www.cs.umbc.edu/agents
URL: AgentWeb News http://www.cs.umbc.edu/agents/agentnews/
- A number of papers discussing the principles behind developing intelligent agents and their application including papers entitled: "The concepts of agent and intelligent agent" and "Abstract intelligent agents", which discusses the use of hypothetical models for defining intelligence in agents.
- Artificial Intelligence Resources
- AIRG multi-agent systems webliography is a collection of potentially useful sources of information about multi-agent systems.
- Agent Sourcebook provides information on agent technology, discussion mailing list and links to corporate and academic research sites.
- "Price-war dynamics in a free-market economy of software agents" is the title of a paper which explores the scenario that in the future: vast numbers of software agents will be providing, trading, and using a rich variety of information goods and services in an open, essentially free-market economy. The paper takes as its starting point that an essential task in such an economy is the retailing or brokering of information, whereby information is gathered from the right producers and distributed to the right consumers. This paper investigates one crucial aspect of brokers' dynamical behaviour, their price-setting mechanisms, in the context of a simple information filtering economy.
- A related piece published in the New Scientist and entitled "Wired for mayhem" considers the potential downside of agent mediated commerce and includes a few links to related resources.
- For a description of JATLite, a free agent platform from the Stanford University CDR.
- The Mining Company publishes a software agents site which includes articles and resource links which are updated every week.
- A mobile agents mailing list for discussion of mobile agents, remote programming, active packets, mobile code, and related ideas. To subscribe, send an email with "subscribe mobility" in the message body. New members also need to submit a two or three sentence biography about their interest in mobile agents.
URL: subscription mailto:firstname.lastname@example.org
- Evaluation of Intelligent Systems (EIS) is an online resource that provides information on the empirical behaviour of intelligent systems. EIS covers exploratory data analysis; hypothesis testing; modeling and statistical terminology and was developed by The Experimental Knowledge Systems Laboratory at the University of Massachusetts.
- The December 1996 edition of IEEE Expert magazine, Vol. 11 No. 6, was largely devoted to articles on Intelligent Agents, whilst the July/August edition of IEEE Internet Computing, Vol. 1 No. 4 was devoted to Internet Agents. Both issues contain articles with extensive lists of web sites covering the subject. One of the papers published in IEEE Expert addressing the central theme of: "Agent-Based Engineering, the Web, and Intelligence" is available online.
- "FAQ for agents" contains a basic introduction to agents and pointers to further information and other resources.
- For a comprehensive listing of research-based agent systems, along with details of standardisation issues visit the The Agent Society home page.
- Intelligent software agents is a site with a large range of information on agent technology and projects, with a good European coverage. The site differentiates between:
Information integration agents - ACL, KIF and KQML
Coordinated agents - cooperative problem-solving, multi-agent systems
Mobile agents - Java, Telescript, Tcl
Assistant agents - personal assistants, softbots, data mining
Believable agents - alife, simulation
- "Agent Standards" paper
- Foundation for Intelligent Physical Agents (FIPA) - to promote the development of specifications of generic agent technologies for interoperability within and across agent-based applications.
- Knowledge Sharing Effort
- IETF AgentX Working Group
- OMG Mobile Agent System Interoperability Facility
A paper entitled: "What's an Agent Anyway? A sociological case study" by Leonard Foner can be found at his site. To show how good a paper it is, here is a short excerpt:
- Autonomy. Any agent should have a measure of autonomy from its user. Otherwise, it's just a glorified front-end, irrevocably fixed, lock-step, to the actions of its user. A more autonomous agent can pursue agenda independently of its user. This requires aspects of periodic action, spontaneous execution, and initiative, in that the agent must be able to take preemptive or independent actions that will eventually benefit the user.
- Personalizability. The point of an agent is to enable people to do some task better. Since people don't all do the same tasks, and even those who share the same task do it in different ways, an agent must be educable in the task in hand and how to do it. Ideally, there should be components of learning (so the user does not necessarily have to program the agent explicitly; certain agents can already learn by looking over the user's shoulder) and memory (so this education is not repeatedly lost).
- Discourse. For all but the simplest of tasks, we generally need to be assured that the agent shares our agenda and can carry out the task the way we want it done. This generally requires a discourse with the agent, a two-way feedback, in which both parties make their intentions and abilities known, and mutually agree on something resembling a contract about what is to be done, and by whom. This discourse may be in the form of a single conversation, or a higher-level discourse in which the user and the agent repeatedly interact, but both parties remember previous interactions.
By this metric, for example, a hammer is not an agent - I don't have a discourse with my hammer! Neither is a Lisp garbage collector, even though it takes spontaneous action to keep my computational environment clean, nor is an automatic transmission in a car; both are autonomous and relatively spontaneous, but it can hardly be said that I have much of a discourse with them.
Booking a flight through a human travel agent, in my case, is only partially a discourse: since I don't have direct access to the actual booking computer, I have no other option. And since I do not have a regular travel agent who knows me, every travel agent is a new experience (albeit, one that travel agencies try to standardize a bit, so as to align everyone's expectations). Now, in one respect, the conversational interchange is a discourse, because, for the duration of the task at hand (booking that one flight), there is a two-way communication of desires and capabilities. However, viewed in a larger context, there is no discourse that 'teaches' the travel agent what my preferences are on when I like to travel, on which airline, and so forth. Viewed this way, a 'travel agent' is nothing more than a somewhat more 'user-friendly' interface to the flight reservation data system.
- Risk and trust. The idea of an agent is intimately tied up with the notion of delegation. We cannot delegate a task to someone or something else if we do not have at least a reasonable assurance that the entity to which we delegated can carry out the task we wanted, to our specifications. However, by definition, delegation implies relinquishing control of a task to an entity with different memories, experiences, and possibly agendas. Thus, by not doing something ourselves, we open ourselves up to a certain risk that the agent will do something wrong. This means that we have to balance the risk that the agent will do something wrong with the trust that it will do it right. This decision must be based on both our internal mental model of what the agent will do (hence how much we trust it) and the domain of interest (hence how much a mistake will cost us).
- Domain. The domain of interest is crucial, as mentioned above when talking about risk and trust. If the domain is a game or a social pursuit, most failures of the agent carry relatively low risk, meaning that we can afford to invest the agent with a considerable degree of trust. On the other hand, the 'fuzziness' and unpredictability of most agent-based systems might make one think twice about using such a system, say, for the control rod feedback system of a nuclear reactor.
- Graceful degradation. Bound up in the notions of risk, trust, and domain, agents work best when they exhibit graceful degradation in cases of a communications mismatch (the two parties do not necessarily communicate well, and may not realize it) or a domain mismatch (one or both parties are simply out of their element, and again may not realize it). If most of a task can still be accomplished, instead of failing to accomplish any of the task, this is generally a better outcome, and gives the user more trust in the agent's performance.
- Cooperation. The user and the agent are essentially collaborating in constructing a contract. The user is specifying what actions should be performed on his or her behalf, and the agent is specifying what it can do and providing results. This is often best viewed as a two-way conversation, in which each party may ask questions of the other to verify that both sides are in agreement about what is going on. As such, the two parties interact more as peers in agent-oriented systems; in non-agent-oriented systems, the user typically 'commands', through some interface, a particular action, and is probably never asked a question about the action unless something goes wrong. In a strictly agent-less situation (eg. a text editor), the feel of the interaction is different than in the case of an agent, primarily due to the discourse-oriented nature of the interaction with the agent and the more 'stimulus-response' or 'non-conversational' feel of the text editor.
- Anthropomorphism. There is a great debate over anthropomorphism in user interface design, and agent-oriented systems would find it difficult to stay out of the morass. However, let me make the point that there are several extent systems that might fit the 'agent' mold that are clearly not anthropomorphic (eg. mail-sorting programs that learn how to sort based on watching the user's actions and making inferences), while there are some that clearly are (eg. Julia). Hence, agency does not necessarily imply a need for anthropomorphism. Conversely, just because a program pretends to be anthropomorphic does not make it an agent: ELIZA pretends to be human, but no one could call it an agent; it lacks most of the crucial notions above, such as being useful for a domain, being personalizable, or having any degree of autonomy.
- Expectations. Whenever one interacts with some other entity, whether that entity is human or cybernetic, the interaction goes better if one's expectations match reality. By the very nature of delegation, assuming perfect operation, especially in a changing world where tasks, goals, and means may be constantly changing, is likely to lead to disappointment. Agents are most useful in domains in which graceful degradation and the correct balance of risk to trust can be obtained, and users' expectations are very important in establishing this domain and making the agent useful.
A thesis entitled "Modelling consumer behaviour", published on the web in .pdf format describes a meta-model of behaviour, which integrates a number of behavioural theories.
This meta-model is being used to develop a multi-agent simulation of consumer behaviour, the so-called "consumat approach". The individual agents are equipped with a number of needs (2, 3 or 4), have different abilities, and may choose between different opportunities to satisfy their needs.
An essential feature of the consumat approach is that the agents may employ 4 types of decision strategies: deliberation (rational actor), social comparison, repetition (habitual behaviour) and imitation, depending on their uncertainty and satisfaction. News, new applications and work in progress concerning the consumat approach can be found at the consumat site.
URL: thesis http://docserver.ub.rug.nl/eldoc/dis/ppsw/w.jager/
URL: consumat http://go.to/consumats/
In a paper entitled: "Domain Specific Languages for ad hoc Distributed Applications", Matthew Fuchs postulates a framework to enable truly "ad hoc cooperative transactions on the Internet, combining human and computational entities together". His paper is based on the premise that the current framework makes a fundamental distinction between human agents (who use HTML) and computational agents, which use CORBA or COM. He proposes Domain Specific Languages (DSLs) as a means to allow all kinds of agents to "speak the same language". The approach adopts ideas (and syntax) from SGML/XML, especially the strict separation of syntax and semantics, so each agent in a collaboration is capable of applying a behavioral semantics appropriate to its role (buyer, seller, editor).
The DARPA Agent Markup Language (DAML) Programme began in the US on 14 August, 2000. The goal of the DAML effort is to develop a language and tools to facilitate the concept of the semantic web through the creation of "technologies that will enable software agents to dynamically identify and understand information sources, and to provide interoperability between agents in a semantic manner".
This goal will be pursued by a research plan that includes the following six tasks:
- Create an Agent Mark-Up Language (DAML) built upon XML that allows users to provide machine-readable semantic annotations for specific communities of interest.
- Create tools that embed DAML markup on to web pages and other information sources in a manner that is transparent and beneficial to the users.
- Use these tools to build up, instantiate, operate, and test sets of agent-based programs that markup and use DAML.
- Measure, via empirical experimentation, the productivity improvements provided by these tools.
- Apply these tools to third party agent development, military-specific problems, and support for the intelligence community so as to evolve DAML technologies towards large-scale use.
- Transition DAML to the commercial and military markets via partnerships with industrial and defense-related (C2 - Command & Control - and intelligence) organisations.
Find out more from the public portal below.
Extempo has received an award of around US$ 2 million from the US Department of Commerce Advanced Technology Program (ATP) for research and development of technology for its Web Learning Guides concept. Web Learning Guides are apparently smart animated characters who "assist learners in web-based learning environments". The research will concentrate on how such technology advances will serve the growing markets for web-based learning systems, including corporate training, secondary and higher education, and independent life-long learning.
Extempo is hoping that its "Web Learning Guides will appear as animated characters who converse with learners one-on-one in natural language, manipulate web content, and adapt their assistance to the learner's needs. They will maintain records of the learner's achievement and provide feedback and encouragement". They will also be developing Guide Authoring Tools to enable educators to author both the subject matter contents and the teaching styles of individual Web Learning Guides. Extempo's proprietary Imp Character Technology will provide the foundation for this project.
This project is one of 37 industrial research projects for cooperative support by NISTs Advanced Technology Program (ATP) 1999 competition, which attracted over 400 proposals. The Advanced Technology Program is managed by the National Institute of Standards and Technology, an agency of the Commerce Department's Technology Administration.
The ATP provides "cost-shared funding to industry for high-risk R&D projects with the potential to spark important, broad-based economic benefits for the US. ATP support significantly accelerates potentially important R&D projects. These are projects that industry on its own could not fully support because of the technical risks involved, and where timing is critical to eventual economic success in the highly competitive global market".
URL: ATP http://www.atp.nist.gov/atp/overview.htm
URL: ATP's 1999 competition http://www.nist.gov/public_affairs/releases/g99-192.htm
URL: Adaptive web learning guides proposal http://www.atp.nist.gov/www/comps/briefs/99013055.htm
URL: Extempo http://www.extempo.com/
The September 2000 News Notes on Agent-based Computational Economics (ACE) have been published online by the Department of Economics, Iowa State University. ACE is "the computational study of economies modelled as evolving systems of autonomous interacting agents".
The news notes contain news items on books, journal announcements, software, research group sites, workshops and meetings, and program, course, and position announcements of possible interest to ACE researchers. Items of more permanent interest are incorporated into the relevant resource sites linked to the main ACE web site.
URL: News Notes http://www.econ.iastate.edu/tesfatsi/ace0900.htm
URL: ACE http://www.econ.iastate.edu/tesfatsi/ace.htm
Mereware educational robotics resources is a site developed to provide "an educational-tech resource for robotic research, artificial intelligence and more". The site features robotics news and also a section which features links to the free web tools used for building the site.
File Downloads - Please note
|File downloads from the El.pub site are currently suspended - the links however have not been updated to reflect this. If you would like access to a particular download file - please email email@example.com with a suitable request confirming a description of the file you wish to download.|
El.pub - Interactive
Electronic Publishing R & D News and Resources
We welcome feedback and contributions to the information service, and proposals for subjects for the news service (mail to: firstname.lastname@example.org)
Edited by: Logical Events Limited - electronic marketing, search engine marketing, pay per click advertising, search engine optimisation, website optimisation consultants in London, UK. Visit our website at: www.logicalevents.org
Last up-dated: 26 March 2021
© 2021 Copyright and disclaimer El.pub and www.elpub.org are brand names owned by Logical Events Limited - no unauthorised use of them or the contents of this website is permitted without prior permission.