Abstract

With the application domain of data mining technologies extends increasingly, some shortages of the commonly used commercial data mining tools, such as poor maintainability, less expansibility and high cost, are gradually exposed in practice. As such, these tools cannot flexibly and effectively meet the apt-to-change demands of users. This paper proposes a plugin style method for dynamic algorithm base design, discusses issues related to the joint between data set and mining algorithms and implements the platform prototype, providing a frame reference for constructing a wieldy, usable, extensible and low cost data mining platform.

This work is licensed under a Creative Commons Attribution 4.0 License.