Knowledge Discovery from Databases (KDD) process is complex, iterative and interactive. It takes place several phases. For its implementation, several modules should be developed (module for data storage, module for processing data, data mining module, evaluation module, knowledge management module). The objective of this study is to propose an approach which assimilates every module to an agent. These agents have to communicate and cooperate to help the user to make the most appropriate decision. Thus, The process of KDD can be likened to a Multi-Agent System (MAS). To validate our approach, we have applied a process of KDD for the fight against nosocomial infections within an intensive care unit (ICU) of a University hospital. On a technical level, we have developed a software tool for decision-making support in Java/XML through the agent platform “Madkit”.
An agent-based Knowledge Discovery from Databases applied in healthcare domain
2013-05-01
1137770 byte
Conference paper
Electronic Resource
English
Expanding the Use of Pavement Condition Data Through Knowledge Discovery in Databases
British Library Conference Proceedings | 2002
|Knowledge-based techniques for product databases
Online Contents | 1999
|Knowledge-based techniques for product databases
Automotive engineering | 1999
|