Distributed AgentsPosted: August 1, 2003
After all the hype behind Artificial Intelligence has died down, things are leaning towards practical applications of the AI especially in the field of Personalization. While it is true that work here could have picked up wind as early as 1999, it is only recently that Distr. Agents and Personalization are drawing attention.
From agents that can organize and prioritize email to agents that collect news and information, from websites that organize themselves into customized portals to mail clients that evolve and learn their users habits, everything is starting to see the light of day.
In terms of collaboration and group work, agents have a tremendous potential in helping organize and build groups, in helping the groups function and parallelize the movement of information between them. In other words agents are becoming conduits to help humans build our own “newspapers” and attract knowledge that might interest us.
Our growing reliance on machines is painted perfectly when we talk about agents. The only difference is that these are’nt tangible aides, they only exist in the abstraction of software.
Agents when combined with Data Mining methods helps leverage the knowledge of algorithms that help abstract Data and at the same time add scalability to the implementation. Agents that mine can be either Sensors on a battlefield or simply “shopping” agents that are looking for the best deal. The objective is the same, to take advantage of the distributed nature of the network and at the same time efficiently consume whatever limited resources are available to return the best possible results.
More links forthcoming…