DMCA
On Compiling Data Mining Tasks to PDDL (2009)
BibTeX
@MISC{Fernández09oncompiling,
author = {Susana Fernández and Fernando Fernández and Alexis Sánchez and Tomás De La Rosa and Javier Ortiz and Daniel Borrajo and David Manzano},
title = {On Compiling Data Mining Tasks to PDDL },
year = {2009}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Data mining is a difficult task that relies on an exploratory and analytic process of large quantities of data in order to discover meaningful patterns and rules. It requires complex methodologies, and the increasing heterogeneity and complexity of available data requires some skills to build the data mining processes, or knowledge flows. The goal of this work is to describe data-mining processes in terms of Automated Planning, which will allow us to automatize the data-mining knowledge flow construction. The work is based on the use of standards both in data mining and automated-planning communities. We use PMML (Predictive Model Markup Language) to describe data mining tasks. From the PMML, a problem description in PDDL can be generated, so any current planning system can be used to generate a plan. This plan is, again, translated to a KFML format (Knowledge Flow file for the WEKA tool), so the plan or data-mining workflow can be executed in WEKA. In this manuscript we describe the languages, how the translation from PMML to PDDL, and from a plan to KFML are performed, and the complete architecture of our system.
Keyphrases
data mining compiling data mining task weka tool knowledge flow data mining task problem description difficult task analytic process data-mining workflow automated-planning community complex methodology data-mining process meaningful pattern current planning system predictive model markup language data-mining knowledge flow construction knowledge flow file large quantity complete architecture available data automated planning kfml format
