Knowledge Management: Reality or Wannabe?

Knowledge management is an emerging category of software that promises to deliver a new way of managing information and enabling companies to better utilize their information technology assets. Given the variety of companies touting knowledge management solutions, however, one must wonder if knowledge management is real or just a buzzword in search of a market. Since knowledge management is in danger of becoming a meaningless term, I'll address three issues in this column: defining knowledge management, how it differs from business intelligence and what companies should do about knowledge management.

Knowledge management is the process of locating, evaluating, capturing and sharing knowledge throughout a company. It relies on establishing a methodology for collecting, evaluating and cataloging information, and making that information available for collaborative access by a wide variety of users and applications within the enterprise.

The goal of a knowledge management solution is to capture the knowledge of people with expertise in specific domains and make that knowledge available for use throughout the enterprise. The biggest problem that knowledge management systems face is extracting the data from people's heads; this is the intellectual capital that has the greatest value to the organization. One way is by scanning various types of text files, such as proposals, e-mails, white papers, transcripts or meeting notes.

Once this information has been appropriately formatted, the knowledge management system figures out the best way to categorize the information. Categorization is essential to the success of a knowledge management solution because it provides the framework for other users to locate information. Some systems enable a human user to define the categories that the information relates to by using keywords or hierarchical list structures. Other systems build a map based on advanced semantic text analysis.

Some knowledge management solutions create a specialized database -- called a knowledgebase -- to facilitate the capturing and categorizing of data. Since knowledge consists of concepts, notions, practices and relationships, the data are not amenable to the two-dimensional structure of the traditional relational data model. As a result, the knowledgebases rely on nonrelational technology, such as object-oriented or text-based databases, e-mail or specialized file structures. Lotus Notes is an example of a popular, proprietary database used for text-based knowledge management systems.

Knowledge management is different from business intelligence in several respects. Knowledge management is based on a set of practices that are implemented in software. Business intelligence tools don't incorporate these practices, which include data categorization and a high degree of embedded collaboration between users. Business intelligence tools use a multidimensional model for analyzing relatively structured data, such as product sales or customer lifetime value. Knowledge management tools focus on analysis of relatively unstructured, text-based data. Business intelligence tools aren't typically part of an organization's workflow, while knowledge management tools are most effective when they are embedded into a company's operational environment.

Knowledge management and business intelligence tools overlap somewhat in how they apply user profiles. Both knowledge management and business intelligence tools can be structured to examine users profiles and give users the ability to subscribe to information of interest. The tool can then use the profile to assess incoming data and notify the user when an item or report of interest is added to the system.

The bottom line for knowledge management is that the concept is nice, but the technology is still relatively immature. The market is fragmented with no clear leader and interoperability standards are lacking. No one has come up with a killer knowledge management application, although a number of companies are developing some innovative knowledge management approaches.

If you work for a services-oriented company, particularly in the high tech industry, then you are probably doing knowledge management manually. Companies that can benefit from the ability to consolidate, categorize and distribute unstructured information to workers should be looking at implementing a knowledge management solution. Recognize, however, that the only viable approach is to implement it yourself, using a tool. Other companies should wait until the market is somewhat more stable, and until true knowledge management applications, not just tools, appear on the market. --Robert Craig is director, Data Warehousing and Business Intelligence Division, at Hurwitz Group Inc. (Framingham, Mass.). Contact him at or via the Web at