Web-Based Data Mining for Quenching Analysis

Document Type

Conference Proceeding

Publication Date

12-1-2003

Abstract

A Web-based Data Mining tool called QuenchMiner™ is being developed for the analysis of quenching data obtained from CHTE at WPI. QuenchMiner is incremental to QuenchPAD™, the existing Database System supporting the CHTE Quench Probe System. QuenchPAD™stores experimental temperature-time data for metal probes quenched in liquid and gas based quenchants. The data is used to compute cooling rates and heat transfer coefficients. QuenchMiner™is a user-friendly tool for querying and analysis. In addition to providing the features of QuenchPAD™ on the Web for worldwide access, QuenchMiner™ does decision making based on case studies and pragmatic knowledge. It utilizes the principles of Data Mining and Knowledge Discovery in Databases (KDD) for this purpose. It uses information from complex data types like graphs, pictures and tables, in addition to simple relational data types. It builds a Knowledge Base of Association Rules and paths for action. It computes suggested decisions in response to the user's scenarios in quenching. QuenchMiner™ provides at-a-glance information for quick and easy analysis, and serves as a Decision Support System (DSS) in heat treating at CHTE. It sets the stage for an Expert System.

This document is currently not available here.

Share

COinS