The research and development of intelligent information agents
is of rapidly increasing importance. In fact, it can be seen as
one of the key technologies for the Internet.
Information agents are computational software systems that
have access to multiple, heterogeneous and geographically
distributed information sources.
One of the main task of such agents is to assist their users
in finding informations they need on the Internet, or,
in other words, to manage and overcome the user's information overload.
Information agents not only have to provide transparent access to many
different information sources in the Internet, but also to be able to
retrieve, analyze, manipulate, and integrate heterogeneous data and
information on demand, preferably in a just-in-time fashion.
Information matchmaking and brokering on the Internet are techniques to enable
collaboration among heterogeneous agents. Both techniques require in
particular
a
common language for the description and automated processing of advertised and
requested capabilities of agents. First steps are taken in this direction,
such as an intelligent use of standardized markup languages, like SGML or XML,
or the more powerful agents capability description language LARKS
(Language for Advertisement and Request for Knowledge Sharing) developed
at the Carnegie Mellon University.
In this talk I present a rough classification of different kinds of
information agents, briefly survey actual systems and implementations,
and show how the LARKS language can be used for information matchmaking
among agents in some detail.