Rita E. Knox, David Newman, Debra Logan
Publication Date: 20 April 2006

Different approaches can help build access to what is labeled "unstructured" content. Architecture work that identifies implicit structure and exploits information access tools, such as search, improve the span of access to an organization's content.

What You Need To Know
Information access can be used to provide information specific to a business's needs. Companies should be sure they aren't missing opportunities enabled by these technologies. Cost justifications are often easy to establish.

Strategic Planning Assumption(s)
By 2007, information architects will establish frameworks to improve the accuracy, accessibility and relevance of enterprise information (0.7 probability).

By year-end 2009, information access will be used as much as content management to support business applications (0.7 probability).

Analysis
Enterprise information covers the spectrum from structured to unstructured content and is used for the full range of enterprise activities – from transactions to collaboration. Databases manage traditional structured data for applications to support, for example, transaction processing, analytics, business intelligence or data warehousing. At the other extreme, unstructured data is often not managed at all. Truly unstructured content (pixels or images) is beyond automated management that recognizes what the content is, who would want to see it or who can change it. Although most Global 2000 companies have implemented some components of content management, only 20 percent have implemented records management, and e-mail archiving is still growing. Some benefits of these technologies do not materialize because of over-reliance on users to classify documents or provide metadata.

The content continuum (see Figure 1) highlights important problems. Widely used, structured data repositories cost money to build and maintain but are used to drive critical business operations, so their value is usually understood. The rigor, discipline, methods, skill sets, experiences and life cycles used to build systems on the structured side are foundational to every successful IT organization. At the unstructured end of the continuum, decision makers struggle to assign value to the content they want to manage. For example, although everyone agrees that contracts are important, there is usually a short list of documents that everyone can point to as valuable. How then do companies decide where to invest in technologies to support human-mediated processes that depend on unstructured content? As organizations attempt to bridge the content continuum and achieve the benefits of integrated content access, lessons learned from the structured realm are considered for addressing issues in the less-structured domain. Contributions come from multiple avenues.

Figure 1. The Content Continuum

Different approaches contribute supporting management across the content continuum. Structuring content to make it machine-processable requires software, hardware, modeling, automated classification and novel approaches.

User Contributions
Content management systems typically require users to enter document metadata to identify parameters that systems can use to control access and organization. Although many applications have become "smarter" about how to automate this process (for example, the author's organization is used to determine who has access), there is still a need for more information to help manage content as desired. Companies often have difficulty getting users to enter requisite data or to do it consistently from one document or user to the next.

Information Architects Information architects can be invaluable in improving the management process. They can identify information-intensive processes (requirements to create, capture, organize, access and use information assets) and create structures (for example, schema) that will help do so. However, if the results aren't applied, then there is little value to be derived. By 2007, information architects will establish frameworks to improve the accuracy, accessibility and relevance of enterprise information (0.7 probability).

Leverage Advanced Information Access Functions
Content structures can be discovered and leveraged in many ways. One is to use information access (search, classification, categorization and taxonomy management; see Note 1) as the engine to identify content components in ways that computers can process them. For example, information access is used to support business processes – finding product information for customers, identifying clients who are candidates for upselling or identifying terms clients use to access online information. It is increasingly integrated as part of enterprise content management suites to support content access, but is also used as a stand-alone technology. Information access includes functions that enable structured viewing of relevant information in ways that are consistent with the kind of information being sought. Examples of information access enabling content use are illustrated in Figure 2. Information access is depicted by the symbol waving hands.

Figure 2. How Information Access Can Leverage Content

Three Examples of Leveraging Information Access

• The first example illustrates CRM use. CRM depends on rapid access to precise information (whether structured or unstructured) that addresses a client's question. Finding the right information will depend on metadata included in content streams. Information access applications can identify or add such metadata when processing content. Successful explanations provide multiple information representations (illustrations with graphics and lists describing actions) delivered in the sequence in which they must be performed. This leads to shorter call time per contact, as well as satisfied customers. The same principles are used to expose this information externally and support customer self-service applications.

 

• The second example shows negotiations arbitrated in an auction. Auction-side support provides product information, along with transaction processing, vendor interaction and registration of client information. User-side support identifies bidders, products sought and account information.

 

• The last example shows information access used to support individual users. Personal search provides access to multiple content data types and stores. This is extremely important because few users have the organizational or taxonomy creation skills to maintain the rigorous categorization that leads to fast information retrieval. Documents, e-mail and instant messages can be retrieved as integrated results.

These examples illustrate the essence of the computing challenge: How do people know what to ask for, and how does a computer figure out what that is? The degree to which the "views" overlap will determine success.

Most companies leverage content – product descriptions, procedures and client information – to drive their business. Not being able to find what is necessary presents significant obstacles to getting work done. In many ways, however, the tools provided by content management applications are not sufficient to deliver what is needed at any point in time. This presents new opportunities for information architects to apply data modeling and database design to all content types. By year-end 2009, information access will be used as much as content management to support business applications (0.7 probability).

Note 1: Information Access Defined
We use the term "information access" to encompass a collection of technologies, including search; content classification, categorization and clustering; fact and entity extraction; taxonomy creation and management; information presentation (for example, visualization) to support analysis and understanding; and desktop (or personal knowledge) search to address user-controlled repositories to locate and invoke documents, data, e-mail and intelligence.

Gartner RAS Core Research Note G00138156, Rita E. Knox, David Newman, Debra Logan, 20 April 2006