Seminar on Data Warehousing

[Home][Presentations][Data Warehousing]

For years, we have been building data warehouseswith a static character. Only periodically (for example every week or every month) new data is added to a data warehouse. However, recently users have begun to ask for data that is almost 100% up-to-date. We call this realtime or active data warehouses. This new requirement has a large impact on the tools that can and must be used, and how data warehouses should be designed.

The first group of users of the data warehouse were the managers who had to make decisions. Nowadays, we also see other user groups, such as customers, suppliers, and dealers, who have access to the warehouse and want to analyze data. We also see non-human users, or in other words, automated processes that make preprogrammed decisions. These processes forms a new group of users.

Working with data warehouses leads to data integration. Other forms of integration also exist, including application integration. In this area, the Service Oriented Architecture (SOA) is strongly rising. This seminar highlights the added value of integrating the SOA with the data warehouse.

Besides working with structured data, more and more users want to be able to analyze unstructured data as well..

New technology keeps being developed, and especially in the field of database technology. This seminar will discuss data warehouse applicances, cubing services, business process engines, streaming database servers, and enterprise service busses.

The popular mashup can be used in a Business Intelligence environments to access external sources directly and to integrate external with internal data.

Overview of the status of data warehousing

Introduction to terminology used

From static to realtime or active data warehouses

Determining an implementation strategy: top down or bottom up

Analyzing a decision process

Selecting a data warehouse architecture: one large central data warehouse, several work group warehouses (data marts), or a virtual data warehouse?

Can Extreme Programming be used for data warehouse projects?

From a classic data warehouse architecture to a virtual warehouse architecture, or the rise of the Data Delivery Platform

Six categories of BI tools: executive reporting, managed query, OLAP, datamining, BAM, and spreadsheets

The return of executive reporting with tools from, among others, Business Objects (SAP), Cognos (IBM), Information Builders, Hyperion (Oracle), Microsoft, and SAS

Market overview of OLAP tools.

The rise of the analytic applications; or "BI out-of-the-box"

The role of the Enterprise Information Portal in a data warehouse

Have open source solutions grown-up?

The mashup as alternative for BI tool

How useful are relational database servers, including DB2, Informix, Netezza, Oracle, SQL Server, Sybase, and Teradata?

Market of data warehouse applicances, including those of DATAllegro (Microsoft), Dataupia, Greenplum, HP, Kickfire, Netezza, Teradata, and Sun

Overview of OLAP technology implemented in relational database servers, such as DB2, Oracle11g, and SQL Server

Special database languages and interfaces: MDX, XML for Analysis, and OLE DB for OLAP

The development of open source database servers

OLAP and data mining functionality in a relational database server

Increasing the flexibility of an information model

The demand versus the supply driven analysis approach

Handling facts and dimensional data

Modelling of historic data: from static to dynamic databases

Are star schema and snowflake designs appropriate for operational BI?

Designing from universal data models

What is the added value of Data Vault?

Normalization or denormalization?

Introducing artificial keys (surrogates) - unique within the database

How to deal with derived or aggregated data?

How and where should data be cleaned up?

How well can data profiling tools spot defected data?

Market overview of data profiling tools

Differences between data cleaning and data profiling tools

Possibilities and impossibilities of data cleaning tools

XML as a language for copying data

Market overviws of ETL tools, including those of Ab Initio, Business Objects (SAP), Cognos (IBM), Informatica, iWay, Microsoft, Oracle, and SAS

What are data mart generators, such as BI-Ready and Kalido?

Aspects of copying: extract, transfer, filter, clean, consolidate, and load

Synchronizing the central data warehouse and the data marts

ETL versus SOA, when to use which tool?

Web services for retrieving external data

The importance of meta data for users

The difference between technical and businesslike meta data

The development of operational meta data

What is a Master Data Management system?

Tools for MDM, including those of IBM, Kalido, Oracle, and SAP

Working with unstructured data: data mining on and analysis of text and image

Web services for importing external data

Data warehouses as information source for the outside world

Introduction

Subjects

1. Introduction

2. The Stages of a Data Warehouse Project

3. Selecting BI Tools

4. Selecting a Database Server

5. Logical Design of a Data Warehouse

6. Physical Design of a Data Warehouse

7. The Quality of Data

8. Copying Data - from ETL to SOA

9. Metab Data and Master Data Management

10. Summary and the Future

Designing the New Generation of Data Warehouses
		[Home][Presentations][Data Warehousing]

		Introduction The world of data warehousing keeps evolving. Technologies have become available that offer new possibilities, such as, data warehouse appliances, mashups, and SOA. The new requirements and demands of users also require that warehouse architectures must be created differently. Users want more accurate reports, or in other words, operational BI is rising. In short, the world of data warehousing is changing. This two-day seminarfocusses on all new developments, insights, ideas, and technologies. A must for every data warehouse specialist. Developing a data warehouse is more than just defining a database structure. Several types of tools must be selected, including OLAP-tools (On Line Analytical Processing), reporting tools, analytic applications, ETL tools (Extract Transform Load), portals, and database servers. In addition, an architecture must be selected: should we set up one large central data warehouse, do we use datamarts, or do we want a virtual data warehouse? We must consider the integration with other systems, such as ERP, CRM, and websites. We must also seriously think about how to handle meta data. The following topics are covered in detail: For years, we have been building data warehouseswith a static character. Only periodically (for example every week or every month) new data is added to a data warehouse. However, recently users have begun to ask for data that is almost 100% up-to-date. We call this realtime or active data warehouses. This new requirement has a large impact on the tools that can and must be used, and how data warehouses should be designed. The first group of users of the data warehouse were the managers who had to make decisions. Nowadays, we also see other user groups, such as customers, suppliers, and dealers, who have access to the warehouse and want to analyze data. We also see non-human users, or in other words, automated processes that make preprogrammed decisions. These processes forms a new group of users. Working with data warehouses leads to data integration. Other forms of integration also exist, including application integration. In this area, the Service Oriented Architecture (SOA) is strongly rising. This seminar highlights the added value of integrating the SOA with the data warehouse. Besides working with structured data, more and more users want to be able to analyze unstructured data as well.. New technology keeps being developed, and especially in the field of database technology. This seminar will discuss data warehouse applicances, cubing services, business process engines, streaming database servers, and enterprise service busses. The popular mashup can be used in a Business Intelligence environments to access external sources directly and to integrate external with internal data. In short, after attending this seminar you are updated on the latest developments in the dynamic field of data warehousing. Subjects 1. Introduction Overview of the status of data warehousing Introduction to terminology used From static to realtime or active data warehouses 2. The Stages of a Data Warehouse Project Determining an implementation strategy: top down or bottom up Analyzing a decision process Selecting a data warehouse architecture: one large central data warehouse, several work group warehouses (data marts), or a virtual data warehouse? Can Extreme Programming be used for data warehouse projects? From a classic data warehouse architecture to a virtual warehouse architecture, or the rise of the Data Delivery Platform 3. Selecting BI Tools Six categories of BI tools: executive reporting, managed query, OLAP, datamining, BAM, and spreadsheets The return of executive reporting with tools from, among others, Business Objects (SAP), Cognos (IBM), Information Builders, Hyperion (Oracle), Microsoft, and SAS Market overview of OLAP tools. The rise of the analytic applications; or "BI out-of-the-box" The role of the Enterprise Information Portal in a data warehouse Have open source solutions grown-up? The mashup as alternative for BI tool 4. Selecting a Database Server How useful are relational database servers, including DB2, Informix, Netezza, Oracle, SQL Server, Sybase, and Teradata? Market of data warehouse applicances, including those of DATAllegro (Microsoft), Dataupia, Greenplum, HP, Kickfire, Netezza, Teradata, and Sun Overview of OLAP technology implemented in relational database servers, such as DB2, Oracle11g, and SQL Server Special database languages and interfaces: MDX, XML for Analysis, and OLE DB for OLAP The development of open source database servers OLAP and data mining functionality in a relational database server 5. Logical Design of a Data Warehouse Increasing the flexibility of an information model The demand versus the supply driven analysis approach Handling facts and dimensional data Modelling of historic data: from static to dynamic databases Are star schema and snowflake designs appropriate for operational BI? Designing from universal data models What is the added value of Data Vault? 6. Physical Design of a Data Warehouse Normalization or denormalization? Introducing artificial keys (surrogates) - unique within the database How to deal with derived or aggregated data? 7. The Quality of Data How and where should data be cleaned up? How well can data profiling tools spot defected data? Market overview of data profiling tools Differences between data cleaning and data profiling tools Possibilities and impossibilities of data cleaning tools 8. Copying Data - from ETL to SOA XML as a language for copying data Market overviws of ETL tools, including those of Ab Initio, Business Objects (SAP), Cognos (IBM), Informatica, iWay, Microsoft, Oracle, and SAS What are data mart generators, such as BI-Ready and Kalido? Aspects of copying: extract, transfer, filter, clean, consolidate, and load Synchronizing the central data warehouse and the data marts ETL versus SOA, when to use which tool? Web services for retrieving external data 9. Metab Data and Master Data Management The importance of meta data for users The difference between technical and businesslike meta data The development of operational meta data What is a Master Data Management system? Tools for MDM, including those of IBM, Kalido, Oracle, and SAP 10. Summary and the Future Working with unstructured data: data mining on and analysis of text and image Web services for importing external data Data warehouses as information source for the outside world [Previous][Up][Next] Copyright (c) 2009 R20/Consultancy B.V.. All rights reserved.