Papel de los End Users en el Análisis de Datos en el Data Warehouse, Apuntes de Sistemática

¡Descarga Papel de los End Users en el Análisis de Datos en el Data Warehouse y más Apuntes en PDF de Sistemática solo en Docsity!

DATA

ACCESS

The Information Delivery Machine

Data

Warehouse

The Web

Data

Sources

End

Users

MOLAP

ROLAP

Data Mart

Data

Mining

 (^) The link between the Data Warehouse and the End Users.  (^) Requires the development of a sound architecture.

Data Access and

Analysis Tools

Business Users & Technical Users Technical Users Business End-Users  (^) Titles:  (^) Planners, Analysts, Managers, Product Developers

 Related Work:

 (^) Market Research, Sales Administration, Business Strategy, Customer Service  (^) Data Analysts, DBAs, Operations Manager, Network Administrator, Application Developers  (^) Business users don’t care how the data ended up in the warehouse, what does it takes to maintain it or what technology was used to get it there.

The Business Users

 “Hands-on” Knowledge Workers.

 (^) First - a business person that runs DSS Tools. Second, a technician.  (^) Provides management with analytical reports.

 Top level Executives - the

ultimate users of the information

that comes from the Data

Warehouse.

 (^) Run the business. Emphasis on competitiveness and profitability  (^) CEO, COO, CFO, CIO, Marketing Vice- Presidents, Corporate Strategists

End Users Work 100 200 300 400 500 600 1st qtr 2nd qtr 3rd qtr 4th qtr 1st qtr 2nd qtr total policies W. H. Inmon, “Building the Data Warehouse” Western Region Southeast Region Northeast Region Central Region New York Massachusetts Connecticut New Jersey New Hampshire introduction of “spring colours” option salesman new incentive program competition's next year’s line promotion Jan Feb Mar Apr May Jun Jul Ago Sep Oct Nov Dec corporate revenues Jan Feb Mar Apr May Jun Jul Ago Sep Oct Nov Dec corporate revenues consumer spending index Correlation Drill-Down Event Mapping Business Trends

OLTP Vs DSS Queries Few Indexes Many Rare Joins Common Normalized DBMS Data Redundancy Denormalized DBMS Rare Derived data and Aggregates Common  (^) Small, pre-defined queries.  (^) Short input and output messages  (^) End users don’t write queries.  (^) Data tends to be focused on current values  (^) Queries are typically unplanned.  (^) Queries return larger answer sets and run longer.  (^) End users issue the queries  (^) Data covers long time spans Complex data structures (3NF databases) Relational DBMS. Multidimensional data structures. OLTP DSS

End User Mindset

I'll tell you what I want, what I really really want So tell me what you want, what you really really want

 " Give me what I say I want, then I can

tell you what I really want "

 To be successful, the end user must be able to

explore the possibilities.

Now... we have a Band and a Song

The Client/Server Model

Query, Reporting And Analysis Tools

Data Warehouse

Desktop

OLAP Servers

Personal

Productivity

Tools

Query and

Reporting

Tools

Analysis

Tools

Analysis Tools

Data Marts

Personal Productivity Tools

Data

Warehouse

Database Server Fat Client

Lotus 1-2-

Excel

 (^) End users familiar and comfortable on how to use them.  (^) Included in Office software suites: Microsoft Office, Lotus SmartSuite, Corel  (^) Provide connectivity to data sources, charting, pivot tables, etc.  (^) Also, statistical packages, and graphics tools.

Data Access and Query Tools - Access via Data Marts

Data

Warehouse

Data Marts

Aggregated Data

Database Server Fat Client

Business Objects

Brio

 (^) Query Tool connects to “single-subject” functional/departmental Data Marts.  (^) Performance is improved by storing aggregated data at the Data Marts level.

OLAP Defined Columns Rows Store Product Time Relational Model OLAP Model Cube Hyper Cube Data Cube Numeric/ Time Series Functions Ratios Moving Averages Cumulative Sums Period to date calculations Variance & Percent Variance Analysis and Modeling Annualization Aggregations Forecasting Procedural calculations What-If Financial Functions Depreciation Growth Rate Net Present Value Rate of Return

MOLAP

Data

Warehouse.

Stores atomic and

summarized data.

MOLAP Tools

Oracle Express

Arbor Essbase

Pilot Lightship

 (^) Warehouse data is pre-aggregated and stored in proprietary data structures, also known as “OLAP cubes”.  (^) Provides extremely good response time for interactive queries.  (^) Load times/Size of the “Cube” is an issue for MOLAP tools.

 Proprietary data structure.

 Pre-calculates as many

outcomes as possible.

Provides

multidimensional view

Reach

through

ROLAP

Data

Warehouse

OLAP Server Fat Client  (^) Fetches multidimensional data from the data warehouse. Stores the data on temporary/persistent cubes.