Discussion Locked

WhatzinaName

0 helpful
2 replies
Closed to new repliesThis discussion is closed. It stays readable, but it takes no new replies.
blog root

2 replies

by kesavan_h

Kimball Vs. Inmon

Article published in DM Direct Newsletter
September 1, 1999 Issue

By Susan Gallas

Numerous data warehouse practitioners have questioned the different approaches to data warehousing as taught by Bill Inmon and Ralph Kimball. Religious wars have erupted between devout followers of each. As the industry has matured, the two philosophies have become much more aligned with each other. However, the language used to describe their methodology leads to questions of compatibility.

Actually, their approaches are quite compatible. It's just a question of how Kimball and Inmon name their structures. Both experts agree that the success of the warehouse/marts depends on effectively gathering the business requirements first. These requirements drive the design of the mart which, in turn, drives the data required in the warehouse. The big bang theory of building the warehouse based on the operational systems has proven excessively time- and resource-consuming with limited return to the business.

Both experts agree that business-user validation of the data mart design ensures that expectations are managed. Once this has occurred, the data analyst researches the most appropriate source system(s) and obtains the model (or designs their own if one does not exist). This model is the starting point for the design of the staging area (or warehouse). Kimball calls it the backroom, and Inmon calls it the data warehouse. This is where the referential integrity rules are applied (via the DBMS or software validation) and transformation of disparate values is performed. Other business rules are applied to the data to come up with a consistent format and derived values. Depending on slowly changing dimension requirements, the entity will be defined differently. Type I: Just overwrite; Type II: We need versioning and dates; and Type III: Add old and new copies of certain important fields. Surrogate keys are also generated here. This is necessary for conforming dimensions. Both Kimball and Inmon recommend extracting from a single source once. Some warehouse consultants may have interpreted Kimball's approach to extract multiple times from the same source but I don't think that was his intention. This is all done in the staging area.

Once the data has been loaded into the staging area, the next process is to load the data marts. All parties agree that the most suitable business structure is the star schema. For data mining the data needs to remain in third normal form so hidden relationships can be discovered. Star schema hides those due to extensive denormalization. The star schema closely resembles an Excel spreadsheet along with the use of pivot table. This is the best presentation layer structure.

Regarding the ODS, Inmon's definition is an integrated, volatile, up-to-the-minute picture of the business. This is a useful structure for one-to-one marketing and customer relations, in addition to other areas where only the most recent transactions are important to the operational business process. Kimball describes the ODS as the structure to store detailed transaction data including history which is a reaction to no detailed data in the warehouse. My impression is that if the warehouse were initially designed to hold only lightly summarized data and no detail transactions this was due to limited computer resources and the fact that most business users do not need that detailed information for their analysis. As the power, speed and storage capacity of computers has increased, the transaction detail history belongs in the warehouse where the users can access that level of detail if need be.

What I have done is to merge the two philosophies to come up with the best design principles for the projects I have worked on. Kimball is an expert when it comes to understanding and documenting star schema design and working with the business users. He has created the concept of slowly changing dimensions of reporting - as is versus as was - and has solved every technical issue that has ever been presented to him and those he has discovered on his own. His strength lies in implementation, and I consider him the front-room expert.

Inmon has pioneered the concept of the data warehouse and decision support. I rely on his expertise for warehouse design, data mining and the closed-loop cycle of information. As warehouse practitioners, it is in our best interest to understand the complexities and subtle differences of the two approaches and apply the aspects of each that fit the problem we are trying to solve. Inmon and Kimball have given us a structured methodology to solve the business problems and it is up to us to judiciously apply these principles.

 

 

by kesavan_h

Ref: OutsourceKing.com

To get the most out of Microsoft Project we need to first understand some common terms because ambiguity can cause confusion later as we move through how to use Microsoft Project.

Ambiguity is perhaps one of the greatest causes of failure of projects. Ambiguity causes scope creep, missed deadlines, run-away costs, under/over utilization of resources, and others. It's your job as project manager to identify and eliminate ambiguities, otherwise your project plan is nothing more than a guess.

Let's first look at what is and what is not a project.  We'll use the definition of project from the Project Management Institute, which is:

A project is a temporary endeavor undertaken to create a unique product or service.

A project is consider temporary since once the projects objectives are met, the project team will break-up and go onto other projects. The goal of a project is to create something new, or unique.

Elements of Project Management

All projects have three basic elements: tasks, resources and time. These are interrelated and any change in one has an effect on the other two. This is one area where Microsoft Project excels. Whenever you make any changes, the affect of those changes will become instantly visible through Microsoft Project's graphical presentation of your project.

Think of tasks as individual pieces of work which need to be done. Some typical tasks may include:

  • Reports to management
  • Pieces of code for an application
  • Project definition documents
  • Any small (or large) item that contributes to reaching the state goal of the project

Resources are anything used to meet the stated goals of the project:

  • People
  • Machinery
  • Money

For time you originally start with an estimate (also know as a guess) on how long it may take to reach the stated goal of the project. As the project progresses, the time estimate becomes more solid as each piece of the project is examined and a more firm estimate as to how long it'll take to produce the individual pieces.

These three pieces are interrelated. If you think of the three as a triangle, in order for the triangle to remain balanced, any change on one side required changes on the other two. For example, if a new task is added to the project, you'll need a  resources to work on this new task, and the new task may (or may not) affect the time side of the triangle .

It is within this area, the management of tasks, resources and time, that Microsoft Project excels and will help you to successfully manage your project.

 

 

This discussion is closed to new replies.