Transactional Agents: Towards a Robust Multi-Agent System (2001) (Lecture Notes in Computer Science #2249)

By: Khaled Nagi

Sign Up Now! Already a Member? Log In
You must be logged into UK education collection to access this title. Learn about membership options, or view our freely available titles.

Synopsis: The term “agent” is one of those catchwords that mean widely differing things to different people. To telecommunications people it is little more than a mobile piece of code that may be executed at any place. At the other extreme, AI people often associate with agents human-like traits such as social behavior. In between, software people view agents as fairly self-contained pieces of software that, at the low end, pretty much act like objects and, at the high end, more or less auto- mously decide when and how to react to stimuli or proactively initiate effects that can be observed from their environment. Software agents are particularly important when it comes to distributed en- ronments. There, much of the communication takes place asynchronously, that is the sequence of events cannot be planned ahead in all detail. Instead, agents are given rules as to how to interpret the current situation and, given a common goal, so that they adjust their response accordingly.