Research on personalized data mining system based on Multi-Agent technology
Meng Zu · Journal of Central South University of Technology(Natural Science) · 2003
A personalized data mining system is designed based on MultiAgent technology, and the basic idea of constructing this system is described first. Then UserAgent is designed with detailed introduction of transition relations from one state to another of this agent and their communication protocol, and a MAS which is called personalized data mining system, is established with UserAgent and MiningAgent. The algorithm, on which the system is mainly based, is given. The system is characterized by the fact that an agent serves as a person in data mining system. The application results show that personalized knowledge can be mined in the system and the knowledge, which is really useful for the user, is given. And the system preferably avoids repeated work and unwanted energy of the user. At the same time, the case of mining of the irrelevant data seldom occurs, thus the system can work efficiently.