Home | Business News | Browse by Publication | X | XML Journal

Leveraging IT assets: integrating XML and relational data.

Article Excerpt
XML, is establishing itself as the standard for exchange of formation across enterprises. However, the technology at allows enterprise-class applications to deal with XML processing is still not clearly formulated. This causes most enterprise customers to implement their own architecture. Additionally, their software implementations try to deal with the same set of basic XML processing questions in different layers of the enterprise, rather than as a whole.

*********

Consider a fictitious company that has been using plain text documents or spreadsheets to report its financials. In the late 1990s, the company is asked by regulatory authorities to report more detailed financial information in a new format. The new data format is XML and the standard is XBRL (Extensible Business Reporting Language), an emerging XML-based open source specification for exchanging and processing financial information.

Despite reluctance due to the high-end costs of technology adoption, the CEO agrees to adopt the standard throughout the enterprise. An IT project is begun to foster the adoption of this particular format as the lingua franca for exchange of financial information between different parts of the company.

An IT ream rolls up their sleeves and gets to work. Before the team can create the necessary IT infrastructure that allows existing applications to communicate with each other using the new data format, it stumbles upon the following questions:

* Should the IT department create a new application that talks to all other applications?

* Will the IT department purchase new and expensive software to make the new format pervasive?

* What happens to legacy systems--older and reliable systems still widely used in key business functions like billing?

* Given that financial reporting is such a key part of the enterprise, how does the company ensure that their software solutions based upon the new format are reliable?

The IT department decides to design an application to generate one XBRL balance sheet for the entire company. As it sets out to develop this application, it needs to consider how to utilize the various databases throughout the enterprise that house important financial data. The obvious candidate step is the storage of XML as a Binary Large Object (BLOB). However, the IT department must examine additional steps and explore the possibility of pushing most, if not all, work into a relational database.

The IT department decides to generate XBRL balance sheets from "raw" data for every division of the company. Raw data could be paper-based data, data in spreadsheets, or data in relational databases.

The XBRL balance sheets generated by each division are further consolidated, analyzed, and verified to form one single XBRL balance sheet for the company. It is then issued to regulatory authorities as a set of XBRL documents.

As any application development team would do, the IT department tries to identify the core requirements of this application. The main requirements are identified as XML consumption from many sources, XML storage in an efficient and reliable manner, searching XML to perform analysis and verification, and transforming XML to generate other XML formats (see Figure...