Towards process-oriented tool support for knowledge discovery in databases

Abstract

Abstract. Knowledge Discovery in Databases KDD is currently a hot topic in industry and academia. Although KDD is now widely accepted as a complex process of many di erent phases, the focus of research b ehind most emerging products is on underlying algorithms and modelling techniques. The main bottleneck for KDD applications is not the lack o f techniques. The challenge is to exploit and combine existing algorithms e ectively, and help the user during all phases of the KDD process. In this paper, we describe the project Citrus which addresses these practically relevant issues. Starting from a commercially a v ailable system, we develop a scaleable, extensible tool inherently based on the view of KDD as an interactive and iterative process. We s k etch the main components of this system, namely an information manager for e ective retrieval of data and results, an execution server for e cient execution, and a process support interface for guiding the user through the process.

Ontology-Driven KDD Process Composition

“…An early work of this kind is [9], where authors suggests a framework for guiding users in breaking up a complex KDD task into a sequence of manageable subtasks, which are then mapped to appropriate Data Mining techniques. Such an approach was exploited in [11], where the user is supported in iteratively refining a KDD skeleton process, until executable techniques are available to solve low-level tasks. To this end, algorithms and data are modeled into an object oriented schema.…”

“…Some research works were also proposed for supporting process composition [8,9,10,11,12]. An early work of this kind is [9], where authors suggests a framework for guiding users in breaking up a complex KDD task into a sequence of manageable subtasks, which are then mapped to appropriate Data Mining techniques.…”

A survey of intelligent assistants for data analysis

“…CITRUS. CITRUS [Engels 1996;Wirth et al 1997] is built as an advisory component for Clementine [Grimmer 1996] -a well-known KDD suite, now part of IBM SPSS Modeler 6 . It contains a case-base of available 'processes' (KD operators) and 'streams' (sequences of processes), which were entered by experts and described with pre-and postconditions.…”

Auto-experimentation of KDD Workflows Based on Ontological Planning

“…It becomes cumbersome to select the right operator as well as to build the right workflow. One of the solutions proposed by several authors is to use AI planning techniques to automatically generate workflows [1,23,7,22]. But current implementations are limited since they support a reduced number of operators (not more than 50) as well as the generated workflows contain not more than 10 operators.…”

“…Opposed to other approaches we use Hierarchical Task Planning (HTN) [11] since hierarchical task decomposition (from CITRUS [22]) and the knowledge available in the DM domain (CRISP-DM standard [6]) can significantly reduce the number of generated workflows. Indeed this limits the number of unwanted workflows but it is still hard for a user to decide which workflows to choose and execute.…”