Application Server Discoverer - Oracle

This component allows for the easy end-user query implementation. In essence, Discoverer is an ad hoc query, reporting, analysis, and web publishing tool. Like Crystal Reports and Business Objects (commercial products that generate SQL queries from a graphical display, making database querying possible for those who do not understand SQL syntax), Discoverer provides a GUI metaphor for the specification of Oracle Database content and display format.

In addition, Discoverer is a business analysis intelligence tool, with interfaces with Oracle Clickstream and the Oracle Database. When using Discoverer, the end user develops workbooks. At a high level, a workbookis a bundle of metadata that includes the following components:

• Tables that participate in the query
• Report formatting for the result set
• Calculations to perform on the data

Once defined, these workbooks allow inexperienced end users to easily create ad hoc reports against the Oracle Database using the Discoverer End-User Layer (EUL) graphical user interface. In addition, Discoverer allows end users to view data at several levels, drilling down to more detail or rolling up to summary level.

As you see in Figure 1-9, there are two main phases in Discoverer usage. First, the Discoverer administrator creates the workbooks by specifying the tables, formatting, and computation rules for any given report. Second (the run-time phase), the end user accesses the EUL and creates customized reports using the Discoverer wizards.

The core of administration for Oracle Discoverer is the development and maintenance of the workbooks and metadata objects. For example, each time an end user runs a report, Discoverer refers to the eul_qpp_statistics metadata table in the infrastructure to produce a time estimate for the report. For more details on the administration and use of Discoverer, see Oracle Discoverer Handbook, by Armstrong-Smith and Armstrong-Smith (McGraw-Hill/Osborne, 2000).

Oracle Forms Server

An evolution of the Oracle SQL*Forms application development tool, the Oracle Forms Server was originally used to render screen display from Oracle content. Enhanced to provide support for HTML, Oracle Forms Server is now used within Application Server 10g to render web pages that include Oracle Database content.

Because the Forms Server is the main engine for rendering web pages, tuning and administration of this component are critical aspects of overall Application Server 10g administration. We will discuss Oracle Forms Server administration and tuning in more detail in Chapter 10.

Application Server Personalization

Analyzing page viewing behavior and creating custom web page content on a busy e-commerce site constitute a formidable computing challenge. To address these issues, Oracle has developed the Oracle Application Server Personalization 10g and the Oracle Data Mining suite. Personalization is extremely sophisticated and relies on internal data about end-users’ web page visits, web page clicks, and referrer statistics. Even more powerful, Personalization allows for the incorporation of external metadata such as customer demographics. It is worthwhile to note that Oracle has several competitors in the web personalization market, notably Blue Martini, Vignette, and Personify.

The goal of Personalization is to accurately identify classes of end users and correlate their behavior with the behavior of other known groups of end users. Using sophisticated multivariate correlation techniques, web page contact can be customized according to predictions about each end user’s preference for web page content. The nature of this analysis is very resource intensive, and almost all large Application Server 10g shops devote large servers exclusively to developing these predictive recommendations.

IT marketing professionals know that it is critical to get the right products onto a custom web page. To be successful, Application Server 10g must be able to accurately predict a user’s propensity to buy a product, based on prior buying and browsing patterns, and buying patterns of like-minded customers (customer profiling). The challenge in developing these predictive models is accurately placing visitors into consumer groups. A consumer groupis a group of customers with similar demographics and buying patterns.

Figure 1-10 shows the process of analyzing demographic information to place visitors into consumer groups. A visitor can be placed into a consumer group in two ways:

• Demographic category (collected from personal information)
• Pattern of page views (collected from referrer URLs)

Once consumer groups have been defined in Personalization, you next start a data mining procedure to correlate the patterns of each consumer group with specific products. The customized HTML personalization is based on data from three sources:

• Known consumer group data -- These groups consist of predetermined summaries of consumer group characteristics.
  
• Weighted rankings of pages viewed -- This is a measure of the popularity of product pages according to each consumer group.
  
• Historical data -- This is historical sales data, correlated by consumer group.

Personalization uses these sophisticated consumer group and data mining component mechanisms to create the web content (Figure 1-11). The administration of Personalization is simplified by using the Personalization GUI, and the Oracle documentation has an excellent discussion of Personalization administration.

This chapter is from Oracle Application Server 10g Administration Handbook, by Garmany and Burleson. (McGraw-Hill/Osborne, 2004, ISBN: 0072229586). Check it out at your favorite bookstore today. Buy this book now.

>>> More Oracle Articles          >>> More By McGraw-Hill/Osborne

blog comments powered by Disqus