Oltp

Improving OLTP Data Quality Using Data Warehouse Mechanisms

Matthias Jarke, Christoph Quix

Informatik V, RWTH Aachen, D-52056 Aachen, Germany

{jarke,quix} @informatik.rwth-aachen.de

Guido Blees, Dirk Lehmann, Gunter Michalk, Stefan Stierl

Team4 Systemhaus GmbH, D-52 134 Herzogenrath, Germany

firstname.lastname@team4.de

Abstract

Research and products for the integration of heterogeneous legacy

source databases in data warehousing have addressed numerous

data quality problems in or between the sources. Such a solution

is marketed by Team4 for the decision support of mobile sales

representatives, using advanced view maintenance and replication

management techniques in an environment based on relational

data warehouse technology and Lotus Notes-based client systems.

However, considering totall information supply chain management,

the capture of poor operational data, to be cleaned later in the

data warehouse, appears sub-optimal. Based on the observation

that decision support clients are often closely linked to operational

data entry, we have addressed the problem of mapping the

data warehouse data quality techniques back to data quality

measures for improving OLI’F’ data. The solution requires a

warehouse-to-OLTP workflow which employs a combination of

view maintenance and view update techniques.

1 Introduction

Team4 is a software house specialised on sales force automation

solutions (SFA) for medium and large enterprises, centering

around customiz,ed data warehouses for typically 50

to more than 1000 users. Customers include, among others,

the chemical giant Bayer and Siemens. The company

was started in 1996 by four people, partially from industry,

partially from research, and has since grown to almost 100

employees. Based on tlhe experience gained in specific customer

solutions, the company has for the last few years also

been developing a line of tools for data warehouse development

and data quality.

The basic product strategy of...