Download PDF version
Introduction
The data is the key asset of a company. Companies want to gather information from the data; therefore, they start BI projects. However, most of BI projects have to deal with problems with data quality. Data quality can be a huge obstacle for a successful BI project. Of course, line of business applications suffer from poor data quality as well. Every company has part of the data which is used everywhere, in every transaction, like customers data, products data and similar. Such data is called master data. People who manage master data are often called data stewards. Processes and activities for maintaining master data are known as Master Data Management (MDM). In this seminar, we are going to discuss MDM problems and solutions and introduce Microsoft tools for MDM, including Master Data Services (MDS).
Outline
Module 1: Master Data Management
For an introduction, we should get basic understanding what master data is. The most important questions – why should you manage master data – are explained. An introduction to the complete Master Data Management (MDM), including processes, tools, and people involved in master data management, is following. Finally, we introduce SQL Server 2008 R2 Master Data Services (MDS) and show where MDS fits in the MDM picture.
Module 2: Data Quality
Data is the key asset of a company. Bad data quality can have a huge business impact. In this module, we are going to learn about data quality issues. We are going to introduce data quality dimensions. The attendees will learn what has the most important impact on data quality, including data model, data governance, security, and performance issues. Data profiling is the process of examining the quality of your data. We are going to show how you can do data profiling with SQL Server 2008 R2 tools. It is not important to cleanse the data only; you have to maintain data quality over time. The reader is going to learn how MDS can help with data quality.
Module 3: Master Data Services
In the second module, attendees are going to learn about key concepts and architecture of Master Data Services through an example. We are going to build a simple MDS application. During this practical work, MDS concepts like MDS Hub, models, including entities, hierarchies and collections, are going to be introduced. We are going to stress the importance of maintaining multiple versions of master data. We are going to use MDS data stewardship portal, namely the Master Data Manager Web application. We are going to touch the problems with importing and exporting data as well.
Module 4: Administering Master Data Services
This module is about administering, i.e. installing, securing, maintaining and monitoring your MDS solution. We start with explaining considerations for MDS’ installation. The reader is going to get understanding of all pre-install considerations, including setup limitations, database and Web application considerations, and multi-lingual MDS applications. We are going to briefly touch how to plan for support of importing and exporting data. A practical step-by-step guide to installation closes this chapter. We will continue with introducing MDS security.
In this chapter, we’ll also discuss the concept of Data Governance and how to deal with it using Master Data Services’ tools. The role of a Data Steward in a company will also be introduced, along with the application of a Workflow to notify for business rules violations in order to maintain high-quality data. Finally, we’ll discuss everything related to Master Data Services’ configuration, maintenance and performance monitoring, taking care also of administrative tasks like Backup and Restore of a MDS solution.
Module 5: Complex Scenarios for Merging Data
This is one of the most complex modules of this advanced seminar. Unfortunately, it is not always easy to maintain master data, especially when you have multiple sources without a common key(s) definition. We are going to introduce different algorithms for approximate string matching. We are going to deal with efficient matching by preventing complete cross joins of different sources. Finally, an algorithm for continuous merging is going to be introduced.