Description of Knowledge Discovery Tools in KDTML

Abstract

Knowledge Discovery in Databases (KDD) is a highly complex process where a lot of data manipulation tools with different characteristics have to be used together to reach the goal of previously unknown, potentially useful information extraction. The design of a KDD process implies the search for suitable tools, the understanding of their scope and proper use, their composition and so on. All these activities can be supported by structured knowledge about the tools. This paper is devoted to presenting KDTML, a Knowledge Discovery Tool Markup Language for the annotation of tool characteristics, like the tool location and execution environment, I/O interface and functionalities. As an example of use of KTDML we discuss the implementation of a wrapping service, which allows to automatically transform a KDD tool written in any imperative language into a web service.