With the increase in the usage of databases in various fields and domains, to overcome the challenges in a centralized data mining environment, more and more databases are distributed in networks. The objective of distributed data mining is to perform data mining operations based on the type and availability of distributed resources. To make a proper choice of a particular DDM system/model, the basic differences between each of them must be understood. This paper produces a survey of some of the DDM systems available. It mainly focusses on the homogeneous DDM models. It discusses methods based on semantic web and grid, multi-agent, mobile agent and i-Analyst. A hybrid method AGrIP is also discussed. A comparative analysis is made considering different key issues of DDM. Each method is described in detail by its method/algorithm.
Inhaltsverzeichnis
- INTRODUCTION
- Classification of DDM Systems
- Heterogeneous Vs. Homogeneous
- Homogeneous DDM systems:
- DDM systems based on Data Mining Agents
- DDM models based on Grid
- Meta-Learning based DDM systems
- Heterogeneous Systems
- DDM models based on CDM
- Methods & Architecture
- Extendible Multi Agent Data mining System
- CAKE
- i-Analyst based DDM
- Multi Agent DDM model using AATP
- Mobile Agent in DMM
- DDM based on Semantic Web and Grid
- AGRIP based DDM
- COMPARISON & ANALYSIS
- Comparative Analysis:
- Different approaches
- Challenges
- CONCLUSION
- REFERENCES
Zielsetzung und Themenschwerpunkte
This paper provides a comprehensive survey of distributed data mining (DDM) systems, focusing on homogeneous models. The objective is to analyze and compare different DDM approaches, highlighting their strengths and weaknesses. The paper aims to provide a clear understanding of the various methods and their key characteristics, enabling researchers and practitioners to make informed decisions about selecting appropriate DDM systems for their specific needs.
- Classification of DDM systems (heterogeneous vs. homogeneous)
- Overview of different DDM methods (agent-based, grid-based, meta-learning, semantic web and grid)
- Comparative analysis of DDM approaches based on various criteria (openness, platform independence, result quality, communication cost, integration method, fault tolerance)
- Discussion of challenges faced by current DDM methods
- Future directions for research in DDM
Zusammenfassung der Kapitel
The paper begins by introducing the concept of distributed data mining and its advantages over centralized data mining. It then delves into the classification of DDM systems, distinguishing between heterogeneous and homogeneous approaches. The paper focuses on homogeneous DDM systems, which are further categorized into systems based on data mining agents, grid computing, meta-learning, and semantic web and grid technologies.
The paper then provides detailed explanations of several prominent homogeneous DDM methods, including EMADS, CAKE, i-Analyst, AATP, mobile agent-based DDM, semantic web and grid-based DDM, and AGRIP. Each method is described in terms of its architecture, key components, and functionalities. The paper also presents a comparative analysis of these methods, considering factors such as openness, platform independence, result quality, communication cost, integration method, and fault tolerance.
The paper concludes by discussing the challenges faced by current DDM methods, such as result quality and efficiency. It also highlights potential future research directions, including the integration of cloud computing with DDM and the exploration of new comparison dimensions for DDM approaches.
Schlüsselwörter
The keywords and focus themes of the text include distributed data mining (DDM), homogeneous DDM systems, data mining agents, grid computing, meta-learning, semantic web, grid, agent-based DDM, comparative analysis, DDM challenges, and future research directions.
- Quote paper
- Swetha Reddy Allam (Author), Kotagiri Santhosh (Author), 2014, Survey on Distributed Data Mining Systems, Munich, GRIN Verlag, https://www.grin.com/document/294717
