NETTING IT OUT

Metadata matters for the future of enterprise-wide content management systems. Metadata, data descriptors that describe the underlying data resources, are no longer the domain of database designers and other technical specialists organizing ERP and related structured systems. Rather, with the advent of the XML revolution, we believe that how companies organize and categorize their unstructured content resources will have far-reaching implications for their abilities to create compelling customer-facing environments.

IT'S ALL ABOUT DIGITAL CONTENT!

An Emerging Trend

We are hearing a new drum beat time and again. Metadata management is one of the hottest technical trends driving the next-generation content management systems and solutions. Such diverse products as Documentum in the enterprise content management space, iManage in the collaborative content management space, and SAQQARA in the product content management space, are all emphasizing their metadata management capabilities--their ability to manage tagged content components within their core repositories.

From one perspective, metadata management is nothing new. For years, both relational and object-oriented database products have separated metadata from data and rapidly run queries against the metadata in order to locate the underlying data. Basic search technologies found in groupware and e-mail systems (such as Lotus Notes and Microsoft Outlook) tag documents with various fields such as author, title, and keywords, and then enable people to retrieve individual items based on these fields. Electronic data interchange (EDI) standards define fixed fields and associated values, enabling disparate systems to automatically process structured documents, such as purchase orders and invoices, without human intervention.

But, in years past, each repository has had its own proprietary metadata management scheme, which works well only within the confines of the specific repository or set of systems. Now there are emerging industry standards--based on XML-related vocabularies and self-describing documents--which can substantially extend the capabilities of our automatic interactions. With the advent of XML as the lingua franca for tagging electronic content within our enterprise and inter-enterprise environments, we can readily associate sets of content-related descriptors (or metadata) with the underlying content (or data) in order to achieve new levels of functionality, insight, and intuition. And, like so many other Internet-related innovations, industry standards have profound implications.

Creating a Customer-Centric Perspective

At the end of the day, digitized content is the currency that is driving the digital age. Whether we are viewing a breaking report on our favorite news site, ordering the components for a new computer system from a structured catalog, viewing an aerial photograph, or listening to our favorite tunes, we are continually consuming digitized information. All of this content is stored in different formats in various repositories.

It turns out that how the content is organized and who is responsible for creating the underlying cataloging criteria are important factors in their own right. Snap-on Tools, for example, maintains a staff of former auto-mechanics within its e-business initiative. As subject-matter experts, these people categorize all of the various tools in terms of what customers can do with them. In effect they define the metadata elements associated with individual tools which, in turn, enables Snap-on to create a customer-centric perspective. These people add the essential knowledge and insights which enables Snap-on to automatically cross-sell and up-sell products and to match particular products against individual customers and their personal profiles.

THREE KINDS OF METADATA

What Is Metadata?

Managing the metadata that adds "intelligence" to our cyber-centric interactions is fast becoming a critical aspect of creating a compelling online experience. Metadata describes the type and structure of specific data elements. Metadata, per se, is not content; rather, it is content about content.

Yet all descriptors are not equal. We must recognize that there are three kinds of metadata--attributes with their associated entities, information taxonomies, and schemas. Each kind of metadata is useful for managing digital content in a different way...