Tools to tame XML content

IBM later this month will demonstrate Project Cinnamon, the company's content management technology designed to ease document management and storage through XML tagging.

Big Blue plans to ship a version of Project Cinnamon this fall, bundling it into DB2 Content Manager. Toward that end, IBM last week shipped an early beta of the software tool to a handful of corporate users and business partners.

Cinnamon was born in IBM's Almaden Research Center and is a tool designed to automatically create mappings among different forms of data. By allowing users to define how an XML document gets mapped into a database such as DB2, the technology makes it easier to store those documents and to manage their content.

The upcoming utility hopes to address one of the thornier problems associated with XML-based development. Although XML serves as a clear standard for how content in a document is defined, the schema or definition of that content can be markedly different from document to document. This makes it impractical to place thousands of different documents in even a single data source and be able to retrieve certain data using a single search engine, an IBM representative said.

Some analysts think the upcoming technology can play a central role in helping corporate users crystallize the implementation of their Web services and SOA (service-oriented architecture) visions.

"Enterprises need to service-enable as much content as they can to make the vision of Web services and SOA real. A big part of that is to track down and capture data and derive a meta-data view of it. Any technologies that can help with this will be highly sought after," said Dana Gardner, senior analyst at The Yankee Group.

Currently, users or administrators putting data into IBM's Content Manager have to modify that data "manually" in order for it to be identified by a search engine. It is one of the design goals of Cinnamon to automate that capability, an IBM representative said.

The graphically oriented tool makes it possible for administrators to take an original document and then map the definition of its data to how a Content Manager-based repository defines that same data.

After the mapping is done, any XML-based document placed in the repository with the same data definitions can be automatically mapped to the repository's definitions and thus can be searched and accessed.

Ipedo, another vendor targeting XML management, last month rolled out Version 3.5 of XIP (XML Intelligence Platform). New in Version 3.5 are XML rules that provide semantic inspection of XML documents and views for comparing the contents of XML documents against business rules.

In addition, a Visual XQuery Builder provides a graphical drag-and-drop interface that can automatically create queries across disparate sources. This distributed XQuery-building capability allows XIP 3.5 to manage XML both in remote databases or file systems and in a local XML store, according to Tim Matthews, vice president of marketing at Ipedo.

"In most environments, [enterprises] are using XML as a unifying information architecture," Matthews said. "So the ability to pull information from a variety of sources and use XML as the model is quite attractive."