Managing Business Information—Content in Context

Most existing Document Management systems overlook a critical information retrieval concept by relying upon document content as the primary search mechanism. The discipline of Records Management and the requisite management of document context holds the key to precise and relevant retrieval.

Since the late 1990s, the widespread use of computer technology and the advent of electronic business have profoundly affected the way in which business information is created, disseminated and managed. Word processing, e-mail and e-commerce technologies have resulted in the deployment of document solutions including imaging, workflow, electronic document management, integrated document management and corporate portals—all to help users draft, disseminate and find information more efficiently.

While the uptake of these technologies has been rapid, the value of these systems has been significantly reduced by an over-reliance on content as the major (and in most cases the only) document management attribute.The use of content for indexing and retrieval is attractive, as it allows an organization to manage information with minimal user intervention and with relatively simple database schemas. There are, however, inherent problems with the approach, problems that become particularly evident as the volume of information increases.

Context vs. Content

While a content index will return individual items that meet specific selection criteria, it precludes the system from returning groups of related documents unless all documents meet the specified content criteria. Many of the items returned may be irrelevant to the user. Further, and perhaps more significantly, many relevant items are not returned. The content-focused approach ignores two equally important document attributes: context and structure.

According to the Canadian Description and Classification of Government Records Working Group, “context and structure is as important and, in some respects even more important than content. A record by itself is of little value. It is only when it is combined with other records to document (or provide evidence of) a given activity that it will assume value. As a result, records must be described in sufficient detail to ensure that the relationships between records...other related records and...related business processes and activities can be understood.” (“Approach to the Description and Classification of Government Records,” www.imforumgi.gc.ca/new_docs)

Similarly, the Australia Records Management Standard states that in order to be meaningful “...the contextual linkages of records must carry necessary information to correctly understand the transactions that created and used them. It should be possible to identify a record within the context of broader business activities and functions. The links between records which document a sequence of activities should be maintained.” (“Australian Standard: Records Management AS 4390.1-1996”, Part 3 pg. 6.)

Unfortunately, automatic classification based on content does not ensure that these important contextual relationships are captured and preserved.

In the common vernacular, to be taken “out of context” indicates an omission of facts or a “twisting” of the truth. Frequently, business information provided without context will be used to make or justify business decisions. It logically follows that decisions based on incomplete information may result in less than optimal business outcomes.

Capturing and recording context is not a simple task. Often “context” may only be apparent to the initial recipient. Capturing this contextual information can be extremely difficult. In order to preserve the context of a document, it is necessary to preserve associated documents and to establish document relationships. In some instances, it may be necessary to record additional contextual metadata on the document profile. The additional effort required to annotate contextual information will often be neglected by users.

Information management systems that preserve the contextual relationships between the otherwise unrelated information objects improve information retrieval. If context can be combined with content during the initial search then the result-set will contain fewer irrelevant entries.

Significantly, once a single, relevant document is found, the user can (by navigating contextual relationships) immediately find all of the related documents with minimal effort.

The importance of context in information management is by no means a new phenomenon. Subject-based file-folders, for example, are fundamentally a means of contextually relating correspondence. Similar grouping mechanisms can be applied to electronic documents to ensure that contextually related items are easily retrieved as a single collection.

Evolution of Context Relationships

The nature of information (and of information in documents specifically) is that a single item will often be relevant in multiple contexts (i.e., people other than the original creator will find value in the information). Further, while the information is usually provided in a particular context, over time the information may become useful in other contexts. The capacity for information system to capture the initial context and support additional context relationships as they evolve will greatly facilitate information management and retrieval of information.

The need to manage information in context appears to have been overlooked by many of today’s document technology solutions. The over-reliance on content as a primary management and information retrieval technique will result in information repositories from which precise and reliable information cannot be extracted. The discipline of records management provides a number of fundamental information management principles that can be effectively applied to the management of digital information