Talk:Intelligent agent

From Wikipedia, the free encyclopedia
Jump to: navigation, search
          This article is of interest to the following WikiProjects:
WikiProject Robotics (Rated Start-class, High-importance)
WikiProject icon Intelligent agent is within the scope of WikiProject Robotics, which aims to build a comprehensive and detailed guide to Robotics on Wikipedia. If you would like to participate, you can choose to edit this article, or visit the project page (Talk), where you can join the project and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 High  This article has been rated as High-importance on the project's importance scale.
 
WikiProject Computer science  
WikiProject icon This article is within the scope of WikiProject Computer science, a collaborative effort to improve the coverage of Computer science related articles on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
 ???  This article has not yet received a rating on the project's quality scale.
 ???  This article has not yet received a rating on the project's importance scale.
 
WikiProject Cognitive science (Rated C-class, Mid-importance)
WikiProject icon This article is within the scope of WikiProject Cognitive science, a collaborative effort to improve the coverage of Cognitive science on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.
 

Comments merged from Agent Environment[edit]

Merge[edit]

Consider merging Intelligent agents into this page. It focuses on software "bots," for the most part, so the layout of this article will need reworked to properly incorporate it. --Schultz.Ryan 03:28, 20 Mar 2005 (UTC)

definitely merge this article into intelligent agents and move to intelligent agent -- the term "artificial intelligence agent" was made up by User:Melody on 10 august 2003, while trying to come to terms with this text from agent: "Agents generally --and including some software agents-- have the ability to learn and to reason, logically or otherwise. They may discover facts and rules about the world or about others and may assess and argue these truths and may alter their ontology". The term is used nowhere in academia, but has spread thru the GFDL corpus (wikipedia mirrors) -- 70.29.131.204 03:38, 19 August 2005 (UTC)
re bots: they're all bots; but in fact, the text could be properly divided into sections for learning agents and non-learning ones. Also, as we approach the technological singularity we learn how to learn -- ie: meta-learning, (which in humans is temporal lobe based ithink) and is a hallmark of sentience. -- 70.29.131.204 04:20, 19 August 2005 (UTC)

Does anyone have any links to people who are doing research in this field?? If so they should be posted also.

Page move[edit]

I suggest that this article be edited in any of the following ways:

  • renamed to agent environment since this is what the greater part of this article is abount
  • merged with software agent since it applies to them in general
  • deleted since as mentioned above this naming is wrong and the topic is covered elsewhere

See intelligent agent --moxon 09:33, 22 September 2005 (UTC)

I would keep it[edit]

I suggest keeping this article separate from intelligent agent. Typically, in agents research, these things are thought separatelly. The features of artificial environemtns mentioned here are equally as important as features as agents, in fact, their are accepted generally as a de facto definitions of properties of artificial environemnts. --User:Cyril2 14:33, 08 May 2005 (UTC)

"agents" / merge[edit]

re merge: Due to the extensive rework required for the suggested merger of the mentioned atricles, I took the libirty to create this article from scratch. As stated in their discussions artificial intelligence agent focuses on "agent environments" and intelligent agents on bot-type software agents. I suggest that these articles be renamed/merged as such. I do however invite the authors of those articles to contribute to this new article where appropriate. Please comment. --moxon 10:36, 20 September 2005 (UTC)

In my experience there are three communities who claim the word "agent":

  • the animat community, who use the word to refer to either robots or animals (though they sometimes will accept VR agents these days)
  • the agent-based modelling people, who are mostly social scientists and use fairly trivial agents to do fairly cool research.
  • the software agent people, who are basically logicians and software engineers, but do use (neat) AI and are also the most organized.

I'm not sure this article is going to go anywhere if it doesn't clearly engage with at least one of those communities. Though there are some other people who use agent as you do, notably Russel & Norvig. That is, sort of as a generic expression for a strong AI system with no particular attachment to a research community.--81.6.234.170 21:38, 2 February 2006 (UTC)

Pls. see my comment there. superseded. Steipe (talk) 22:24, 23 April 2009 (UTC)

Different types of intelligent agents not related[edit]

If it is interactive and adaptive, IMO it is an IA --moxon 12:26, 16 March 2007 (UTC)

References[edit]

The references on this page need a serious review! The first reference (Kasavbov) does not contain any of the information stated on the page. The relevant information might be somewhere else in the journal but it is not in the Introduction (which is referenced). Further there are hyperlinks to references 3 and 4 that take you nowhere since references 3 and 4 don´t exist. --Shadowselfs 07:34, 25 January 2008 (UTC)

This should be divided into two articles[edit]

There should be (at least) two separed articles:"Intelligent agent" and "Intelligent agent(computer science)" Current article is confusing and if no-one objects I will create the second article and move half of the text there.82.130.8.142 (talk) 19:11, 26 February 2008 (UTC)

I agree. I think this article should be a disambiguation page, and I think there should be two articles, should be called:
Here's my new idea. Let this article be about intelligent agents or rational agents as defined by Russell & Norvig (merging in agent environment). Cover "intelligent software agents" in software agent. ---- CharlesGillingham (talk) 18:28, 13 August 2008 (UTC)
Yes check.svg Done ---- CharlesGillingham (talk) 16:41, 5 October 2008 (UTC)

Merge complete[edit]

I am hoping to merge rational agent, intelligent agent and agent environment into a single article, since they all seem to be based on chapter 2 of Russell, Stuart J.; Norvig, Peter (2003), Artificial Intelligence: A Modern Approach (2nd ed.), Upper Saddle River, New Jersey: Prentice Hall, ISBN 0-13-790395-2 .

This would create a nice medium-sized article on a single topic from a major source. (So we can be sure how the topic is defined, that it is notable, and that the material is both relevant and verified). Other related ideas (such as software agents with artificial intelligence, or agent (economics) can be covered in other articles, and the intro of intelligent agent will guide you there.

It seems this merge has a long history. Hopefully this will sort it out. ---- CharlesGillingham (talk) 17:04, 13 August 2008 (UTC)

As for Rational agent, see my notes at Talk:Rational agent ---- CharlesGillingham (talk) 17:43, 13 August 2008 (UTC)

Yes check.svg Done ---- CharlesGillingham (talk) 16:41, 5 October 2008 (UTC)

Intelligent agents should not be part (in the Scope of?) of Robotics[edit]

The research on Intelligent Agents is not a subpart of robotics, althought some work of robotic is inspired by the work on Intelligent Agents. Actually, research on Intelligent Agents can be rather be considered as a sub-part of the research conducted on Software agents.

Exemple of group conducting research in this area:

Other terms to be added:

  • believable agents
  • emboddied agents (or interface agents).
    • Note: only how the agents appear to the users via artificial characters
  • cognitive agents
  • reactive agents —Preceding unsigned comment added by Nabeth (talkcontribs) 15:20, 14 May 2008 (UTC)

Other reference:

Nabeth (talk) 15:12, 14 May 2008 (UTC)

There are several definitions of "intelligent agent" being used, I think. The first paragraph of this article is based on Russell and Norvig (2003), the standard AI textbook. The next couple of paragraphs attempt (poorly, I admit) to grant some space to some of the other definitions in use. The rest of the article I can't vouch for. This is an issue with other articles as well, such as software agent, etc.
As I just posted over at software agent, I think what is needed is some more authoritative sources that give an overview of the uses of the term "intelligent agent" (and autonomous agent, software agent, etc). I think CS textbooks are the best source, since it is their job to be general. I think the publications of particular institutions or researchers are a poor source, since they tend to enforce a particular definition, one that might not be used by everyone. ---- CharlesGillingham (talk) 23:54, 6 February 2009 (UTC)
I went through the rest of the article and added footnotes to tie each section derived from Russell and Norvig to their textbook. The text from other sources seems very poor and scattershot to me and has been placed in separate sections of the article. Hopefully an editor with a more comprehensive knowledge of this subject can improve these sections. ---- CharlesGillingham (talk) 22:42, 8 February 2009 (UTC)

Simple intelligent agent image[edit]

Um that simple intelligent agent image makes no sense. How can you have a if then condition which is unlinked to the original perception still influence the action? Shouldn't you have sensor "what is the environment like now" go into the agent, then the agent considers "if then" conditions then submit that to the "action to be taken" then to the "actuators" who send it back to the environment and the system? Sorry i don't normally take time doing anything in wikipedia except for reading it so i don't know how to edit the discussion area properly.. Thanks, skythra —Preceding unsigned comment added by 130.102.158.15 (talk) 04:07, 22 March 2010 (UTC)
I think what is meant is that the decision taken by the box 'action to be done' uses the information from the boxes that point towards it, so 'what the world is like right now' and 'condition-action (if-then) rules'. In that sense, the rules do use information of the perceptions that have been done. - Simeon (talk) 06:25, 20 May 2010 (UTC)

Removed content[edit]

I spotted a major removal of content a while ago: [1] Is it a good idea to reinsert it? Mikael Häggström (talk) 16:16, 29 January 2012 (UTC)