Defining like-minded agents with the aid of visualization
Proceedings of the first international joint conference on Autonomous agents and multiagent systems part 3 - AAMAS '02
Profile carrying agents offer the opportunity of meeting like minds and increasing efficiency in many information search applications. Profiles can also increase the sophistication of relationships and interactions in multi-agent systems in general. Such profiles (feature lists) may be of the agent owner (interests) or of the information sought (specifications). At the same time there is a continuing increase of profiling data of increasing complexity becoming available. The paper first
... s various possibilities for defining profiles and selecting similarity metrics in the agent interaction context. General and domain-specific data sets are specially constructed and used to provide a concrete view of the behaviour of the metrics with visualization. Visualization usually results in xy (or xyz) coordinates for each agent, placing them in a profile space. The use of these coordinates as a means of carrying and comparing one's profile without revealing it to other agents is proposed. Thus this paper extends our previous work on visualizing multivariate data and proximity data in the agent scenario in two ways where visualization is not the end product: using visualization tools in the phase of similarity metric choice; proposing the use of coordinates as a profile. The paper seeks to illustrate, with these applications, an objective of visual data mining, namely the increased integration of visualization and analytic techniques.