FlexibleQL: a flexible and extensible query language for data mining The emerging data mining tools and systems lead naturally to the demand of a powerful data mining language. Recently a few works have been done that resulted at the development of a number of data mining languages but it seems that the following abilities have not been considered as goals in those languages: 1) Flexibility of the language, i.e. the existing languages support a set of predefined algorithms and therefore enforce the use of those predefined algorithms. 2) Extensibility of the language, i.e. the language can not be extended in a way that new data mining features and algorithms are used to answer new queries to satisfy organizational demands. 3) The ability to support different types of data mining algorithms, i.e. most of these languages have been designed just to mine association rules and mining other patterns of data mining, such as classification and clustering, the complete support of decision trees and decision rule mining don’t seem to be the goal of the designers of these languages. 4) Post-processing of the discovered patterns for later use and preprocessing of the data to be mined seem to be weak in some of these languages. In this paper we have tried to consider the following points in the design of our flexible mining query language named FlexibleQL. We show that FlexibleQL is flexible in the scene that users can select and configure their desired data mining process. Furthermore, FlexibleQL is extensible and the administrator of the language system can extend the language so that new mining algorithms and tools can be used in order to answer more inductive queries. Hence, new organizational and managerial demands can be satisfied. Besides FlexibleQL has the ability of data sampling that enables users to run data mining algorithms on a randomly selected parts of a large volume data instead of all the data to increase the run time performance of the mining process. Also FlexibleQL saves the results of data mining in special format which is available for the users (system administrator) and has more ability in post processing of the mined rules for later use.
References in zbMATH (referenced in 1 article , 1 standard article )
Showing result 1 of 1.
- Salajegheh, Afshin; Reza, Mohammad: FlexibleQL: a flexible and extensible query language for data mining (2008)