FIT5095 Data Warehousing

By Dr Rob Meredith

About this podcast   English    United States

Designing, managing and operating data warehouses.
June 5, 2013
This week's lecture is a review of the material we've covered this semester as well as a look at current trends and issues in the data warehousing industry.  We also look at the final exam.
May 24, 2013
IT governance is the means by which organisations determine the allocation of decision-making rights and responsibilities for the management of information technology in an organisation.  This lecture looks at how data warehouses may require a different approach to governance to other kinds of IT systems.  The lecture looks at IT governance in general as well as the academic literature on data warehousing governance specifically.  The argument is made that data warehouses are "chaotic and subversive" systems and as such require a governance 'sandbox' separate to normal IT governance operations.
May 17, 2013
This week lecture looks at the technical decisions needed with regard to platform and vendor selection.  The lecture also examines the idea that data warehouses are chaotic systems and looks at the implications of chaos theory for data warehouse management.
May 10, 2013
This week's lecture looks at the problem of data quality for data warehousing projects.  We look at how data quality is much more than just the data being correct - quality depends on the purpose for which the data is being used, and that varies from user to user.  It also varies from source system to source system to data warehouse system.  Data quality problems are therefore very difficult to fix and can be a major stumbling block for many projects.We also take a quick look at the use of metadata in data warehousing, both technically-oriented metadata for facilitating the flow of data from source to data warehouse user, as well as user-generated metadata.
May 1, 2013
This week we look at some of the organisational issues that can cause problems for data warehousing projects, including Davenport's idea of 'information politics'.  We relate Davenport et al's types of information political structures to our discussion last week of data warehousing architectures.  We then look at several case studies of successful data warehousing projects to see that despite the various 'rules' about how to develop data warehouses discussed in the unit, successful projects can result even when these rules are broken.  Context is important, and maintaining a focus on delivering decision support rather than being caught up with data management is more important than, say, what modelling approach is used to design the warehouse schema.
April 26, 2013
This week we again address two distinct topics.  In the first half of the lecture we look at various architectural approaches to structuring data warehouses: the enterprise-wide data warehouse approach, as well as various 'data mart' approaches including Kimball's "federated bus" architecture.In the second half of the lecture we look at the process of extracting, transforming and loading data into the warehouse from source systems, including some of the jargon and tools available to help developers.
April 18, 2013
This lecture covers two disparate topics in the syllabus.  The first looks at the use of data mining technology.  The second looks at approaches to managing data warehousing project lifecycles - alternatives to the traditional waterfall model of the SDLC that take into account the evolutionary nature of data warehouse systems.
April 10, 2013
The intent of this week's lecture is to highlight the pragmatic nature of designing data warehouses: ideological adherence to one approach or another means that when design requirements shift, less than optimal designs result.  To illustrate this, we look at designing for a common application of business intelligence - understanding customers better as part of an organisation's customer relationship management strategy.  We look at two approaches to designing customer-centric data warehouses from Todman and Kimball.
April 4, 2013
This week we look at some physical design issues related to dimensional modelling including slowly changing dimensional data and hierarchies and aggregates.
March 19, 2013
This week's lecture introduces the basics of dimensional modelling, an alternative approach to designing data warehouse database schemas that address the key design requirements of a simple, intuitive design to encourage end-user exploration and use, and efficient retrieval of data from the warehouse for reporting.

Podcasts like "FIT5095 Data Warehousing"   ·   View all

By Realmworx
By Jamie Moffett Media Design & Production, Inc.
Disclaimer: The podcast and artwork embedded on this page are from Dr Rob Meredith, which is the property of its owner and not affiliated with or endorsed by Listen Notes, Inc.