The title of this workshop (Intelligent Agents) will almost certainly mean different things to different people; some will interpret it as being a forum of discussion of issues in multi-agent systems, while others will consider it the appropriate place to present the latest results of intelligent user-interface research or mobile agent technology. The term agent is troublesome, simply because it is used by so many different communities within Computer Science. Each community has its own unique definition of the term; however, one characteristic which seems to be common is the concept of agents as autonomous entities. In essence, my thesis is that to be truly autonomous, agents must employ some form of learning capability, i.e. that intelligent agents are learning agents. By learning agent, I mean any agent which learns about the environment in which it operates, the user(s) it interacts with, or other agents. In the next section I present a number of scenarios to illustrate this point.
|Title of host publication||Proceedings of the UK Intelligent Agents Workshop|
|Editors||J L Nealon, N S Taylor|
|Publication status||Published - 1997|
- machine learning