Towards Advancing Distributed Data Mining: Intelligent Agent Systems
Mais Haj Qasem, Mohammad Aljaidi, Ghassan Samara, Ayoub Alsarhan, Raed Alazaidah, Yousef Ali Al-Gumaei, Esra F. Alzaghoul · 2023
Distributed data mining (DDM) deals with the analysis of distributed data and proposes algorithmic solutions to perform different data analysis and mining operations in a distributed manner by considering resource constraints. Here, patterns are discovered, and predictions are implemented based on multiple distributed data sources. However, DDM faces several problems in terms of performance and implementation. Specifically, the main problems include issues of autonomy and privacy, which this paper aims to solve. The distributed sites must be in the same environment or organization and under the control of the same administration, implying their agreement on the same classification algorithm. Consequently, the best solution is the use of intelligent agent systems that contain a group of autonomous agents with communication and coordination facilities with different classification algorithms and environments, can collaborate with each other, and can decide whether and when to request information from other agents.