|
|||||||||||||||||||||||
|
|||||||||||||||||||||||
ENTERPRISE-WIDE KNOWLEDGE MANAGEMENT SYSTEMS
Managers and firms must deal with many different kinds of knowledge and knowledge issues. There are three major categories of enterprise-wide knowledge management systems for dealing with these different kinds of knowledge. Some knowledge exists already somewhere in the firm in the form of structured text documents and reports or presentations, and the central problem is organizing this existing structured knowledge into a library and making it accessible throughout the firm. We will call this type of knowledge structured knowledge, and we can refer to these types of systems as structured knowledge systems. The first knowledge management systems that appeared in the late 1980s and 1990s focused on this type of knowledge. Managers may also need information that may exist somewhere inside the firm in the form of less-structured documents, such as e-mail, voice mail, chat room exchanges, videos, digital pictures, brochures, or bulletin boards. We can call this knowledge semistructured knowledge, and we can refer to the systems that focus on this type of knowledge as semistructured knowledge systems (the industry name is digital asset management systems). Systems for structured and semistructured knowledge function as knowledge repositories. A knowledge repository is a collection of internal and external knowledge in a single location for more efficient management and utilization by the organization. Knowledge repositories provide access through enterprise portals and search engine technology and may include tools for accessing information from corporate databases. In still other cases there are no formal or digital documents of any kind, and the knowledge resides in the heads of experienced employees somewhere in the company. Much of this knowledge is tacit knowledge and is rarely written down. Here, the problem faced by managers is building a network that connects knowledge demand with knowledge supply. Knowledge network systems, also known as expertise location and management systems, attempt to perform this function. Knowledge network systems provide an online directory of corporate experts in well-defined knowledge domains and use communication technologies to make it easy for employees to find the appropriate expert in a company. Some knowledge network systems go further by systematizing the solutions being developed by experts and then storing the solutions in a knowledge database as a best practices or frequently asked questions (FAQ) repository. Table 12-2 compares the major categories of enterprise-wide knowledge management systems.
Structured Knowledge Systems Structured knowledge is explicit knowledge that exists in formal documents, as well as in formal rules that organizations derive by observing experts and their decision-making behaviors. The essential problems of structured knowledge are the creation of an appropriate database schema that can collect and organize the information into meaningful categories and the creation of a database that can be easily accessed by employees in a variety of situations. Once the schema is created, each document needs to be “tagged,” or coded, so that it can be retrieved by search engines. Structured knowledge systems perform the function of implementing the coding schema, interfacing with corporate databases where the documents are stored, and creating an enterprise portal environment for employees to use when searching for corporate knowledge. All the major accounting and consulting firms have developed structured document and engagement-based (case-based) repositories of reports from consultants who are working with particular clients. The reports typically are created after the consulting engagement is completed and include detailed descriptions of the consulting objective, participants, and the practices used to achieve the client’s objectives. These reports are placed in a massive database to be used later for training new consultants in the company’s best practices and for preparing new consultants joining an existing on-site consulting team. Accounting firms, for instance, have created large tax law accounting databases that store information on tax policy, the application of that policy to specific client companies, and the advice of in-house tax experts on how local laws work. Perhaps one of the largest private-sector structured knowledge repositories is KPMG’s KWorld. KPMG International is an international tax and accounting firm with 95,000 professionals serving clients through 1,100 offices in 820 cities and 150 countries (KPMG, 2003a). With such a large global base of employees and clients, KPMG faced a number of problems in sharing knowledge, preventing the loss of knowledge as consultants retired or left the firm, disseminating best practices, and coping with information overload of individual consultants. In 1995, KPMG began developing a Web-based knowledge environment known as “Knowledge Web,” or KWeb. KWeb contained databases organized around internal and external knowledge domains of interest to its consultants and partners. In 1999, KPMG rolled out an extension of KWeb called KWorld, which features an integrated set of knowledge content and collaboration tools that can be used worldwide. Figure 12-5 provides an abstract overview of the complex nature of knowledge stored by KWorld. FIGURE 12-5
KWorld’s knowledge domains KWorld is an online environment for gathering, sharing, and managing knowledge. Although it is primarily a document repository, KWorld also provides online collaboration capabilities for the firm’s consultants and an internal reporting system. KWorld stores white papers, presentations, best practice proposals, articles, presentations, internal discussions, marketing materials, engagement histories, news feeds, external industry research, and other intellectual capital. The content is organized into nine levels by KPMG products and market segments. Within each of these domains are many subcategories of knowledge. For instance, the client knowledge domain includes entries on financials, industry dynamics, change dynamics, client organization, client products and customers, and KPMG’s history of engagements (KPMG, 2003). Consultants use KWorld to coordinate their work as a team with a client, and the client is allowed access to the collaboration environment as well. Figure 12-6 illustrates the system processes used by KPMG to capture and organize content for this system.
KPMG has invested heavily in organizational and management capital required to make use of the millions of documents stored in KWorld. KPMG has created a division of knowledge management, headed by a chief knowledge officer. An extensive staff of analysts and librarians assesses the quality of incoming information, ensures its proper categorization, assesses its value, and provides some analysis of its importance. Another example of a structured knowledge system is Roche Labs’ Global Healthcare Intelligence Platform, which integrates documents from multiple sources to provide its professional services group with up-to-date information and expertise relating to new Hoffman-La Roche pharmaceutical products. The system gathers relevant information from global news sources, specialty publishers, health care Web sites, government sources, and the firm’s proprietary internal information systems, indexing, organizing, linking, and updating the information as it moves through the system. Users can search multiple sources and drill down through layers of detail to see relationships among pieces of data. Semistructured Knowledge Systems Semistructured information is all the digital information in a firm that does not exist in a formal document or a formal report that was written by a designated author. It has been estimated that at least 80 percent of an organization’s business content is unstructured—information in folders, messages, memos, proposals, e-mails, graphics, electronic slide presentations, and even videos created in different formats and stored in many locations. In many cases firms have no idea what semistructured content has been created, where is stored, or who is responsible for it. It might be convenient or cost effective not to know this information, but increasingly this is not legal or effective under the influence of laws such as the Sarbanes-Oxley Act of 2002, which requires financial services firms to keep records on the origins of all material corporate documents, including e-mails, brochures, and presentations, and strict internal controls governing their storage. The health care industry in the United States has been similarly impacted by the Health Insurance Portability and Accountability Act (HIPAA) of 1996, which requires health care providers and insurers to track the flow of personal health information meticulously. Firms such as Coca-Cola need to keep track of all the images of the Coca-Cola brand that have been created in the past at all their worldwide offices both to avoid duplicating efforts and to avoid variation from a standard brand image. Without the appropriate tools, a firm may have thousands of content centers located in offices around the world where employees often know only about the materials they create themselves. Searching across departments and offices—from corporate legal to technical engineering to customer correspondence to records, or from New York to London to Tokyo—can be a daunting, if not impossible, task. As a result, semistructured documents are continually reproduced and re-created, knowledge is forever lost, and costs climb ever higher. The essential problem faced by managers is building a database and technical infrastructure that can collect such semistructured information and organize it in a coherent fashion. A number of vendors have responded to this need with systems that can track, store, and organize semistructured documents as well as more structured traditional documents. One of the largest players in this marketplace is Hummingbird, a Canadian software company specializing in “integrated knowledge management systems” (see Figure 12-7). In addition to providing document management in the form of centralized repositories, Hummingbird’s Business Intelligence module pulls data from the firm’s enterprise systems and makes it available firmwide through a portal. A more recent addition to the product line is a rules-based e-mail management program that automatically profiles incoming and outgoing mail messages using rules developed by line managers.
One user of Hummingbird’s enterprise knowledge management system is Hennigan, Bennett and Dorman LLP, a Los Angeles-based law firm. The firm handles many big-name, multiparticipant lawsuits, such as government bankruptcies or institutional shareholder fraud cases. New requests for electronic discovery require a legal team to sift through thousands of electronic messages to search for potential evidence. The firm was inundated with backup tapes of hundreds of thousands of e-mail messages and needed a way to filter that information into something its attorneys could use more easily. The firm implemented Hummingbird’s Enterprise document management system, which automates the capture, manipulation, and distribution of document-based knowledge embedded in e-mail. Instead of sifting through piles of printed copies of e-mails, attorneys can run powerful electronic searches, locating only the e-mails they need for a case and marking them up electronically. The system can also re-create all of the threads of an entire e-mail discussion for attorneys to follow and scan e-mail attachments. Using this system has cut the time to process e-mail in half (Hummingbird, 2003a). Another Hummingbird user is Cuatrecasas, a leading Spanish law firm described in the Window on Management. Cuatrecasas implemented Hummingbird Enterprise to provide a standard platform for organizing and managing both structured and semistructured information. Cuatrecasas maintains many offices in many different locations, each previously with its own information systems. The only way information could be shared among offices was by e-mail, and that was also difficult to manage because of version control problems. These problems were solved by adopting a single platform for enterprisewide knowledge management. ORGANIZING KNOWLEDGE: TAXONOMIES AND TAGGING One of the first challenges that firms face when building knowledge repositories of any kind is the problem of identifying the correct categories to use when classifying documents. It is, of course, possible simply to “dump” millions of documents into a large database and rely on search engine technology to produce results for users. However, a brute search engine approach produces far too many results for the user to cope with and evaluate. Firms are increasingly using a combination of internally developed taxonomies and search engine techniques. A taxonomy is a scheme for classifying information and knowledge in such a way that it can be easily accessed. A taxonomy is like a table of contents in a book or like the Library of Congress system for classifying books and periodicals according to subject matter and author. A business firm can access information much more easily if it devises its own taxonomy for classifying information into logical categories. The more precise the taxonomy, the more relevant are the search results produced by search engines. Once a knowledge taxonomy is produced, documents are all tagged with the proper classification. Generally, Extensible Markup Language (XML) tags are used for this purpose so the documents can be easily retrieved in a Web-based system. Products such as ActiveKnowledge (Autonomy Corporation) and Taxonomy (Semio Corporation) attempt to reduce the burden on users by categorizing documents using an existing corporate taxonomy. Such products consider the user’s prior searches, the context of the search term in the document (the relationships between words in a document), related concepts the user may not have entered, as well as keyword frequency and the popularity of the document. The purpose of these newer tools is to increase the probability that the correct response will be in the first 10 results. Several tools perform auto tagging and reduce the need for managers to develop their own unique taxonomies. Semio’s Tagger software is a categorization and indexing engine that identifies key phrases in documents, assigns relevance factors to these phrases, and organizes the documents into categories, creating XML-based document tags using rules that users can see and modify. Tagger can access more than 200 different document types stored in legacy, enterprise, or other intranet databases. Users can integrate existing taxonomy categories and add, delete, or merge categories after examining how the system responds. Semio claims that its semiautomatic system can achieve 95 percent of the accuracy obtained by manually reviewing and tagging documents in a fraction of the time required for manual efforts (www.semio.com). One user of Semio’s auto-tagging tools is Stanford University’s HighWire Press, which publishes 298 online journals containing more than 12 million articles. When the company expanded its collection in 2001 from 1 million to 12 million articles, it needed a way to automate and expand its indexing process. It also needed to provide researchers with better browsing and searching capabilities to support the discovery of unexpected relationships, to link articles from a variety of disciplines, to identify concepts in articles, and to link these concepts in logical categories. Currently, the system has developed 22,000 categories and more than 300,000 concepts. The system supports 84 million hits each week with a database of 6 terabytes (Semio, 2003). The system requires some active management. HighWire Press reviews its classification scheme every month and makes changes based on user feedback and management insight. Knowledge Network Systems Knowledge network systems address the problem that arises when the appropriate knowledge is not in the form of a digital document but instead resides in the memory of expert individuals in the firm. According to a survey by KPMG, 63 percent of employees in Fortune 500 firms complain of the difficulty in accessing undocumented knowledge as a major problem. Because the knowledge cannot be conveniently found, employees expend significant resources rediscovering knowledge. An International Data Corporation (IDC) study estimated that the average cost of redundant effort in Fortune 500 companies exceeds $60 million per year per firm (AskMe, 2003a). Figure 12-8 illustrates the problem of “collective ignorance,” a situation in which someone in a firm knows the answer, but that knowledge is not collectively shared.
Knowledge network systems seek to turn tacit, unstructured, and undocumented knowledge into explicit knowledge that can be stored in a database. Solutions that are developed by experts and others in the firm are added to the knowledge database. This new knowledge can be stored as recommended best business practices or as an answer in a database of frequently asked questions. Table 12-3 lists some of the key features of enterprise knowledge network systems.
AskMe, Inc., produces a widely adopted enterprise knowledge network system. Its users include Procter & Gamble and Intec Engineering Partnership, a project management company with more than 500 employees worldwide serving the global oil and gas industry. The software, AskMe Enterprise, enables firms to develop a database of employee expertise and know-how, documents, best practices, and FAQs, and then to share that information across the firm using whichever portal technology the firm has adopted. Figure 12-9 illustrates how AskMe Enterprise works. An Intec engineer with a question, for instance, could access relevant documents, Web links, and answers to previous related questions by initiating a keyword search. If the answer could not be found, that person could post a general question on a Web page for categories such as Pipeline or Subsea for other engineers accessing that page to answer. Alternatively, the person could review the profiles of all company engineers with relevant expertise and send a detailed e-mail query to experts who might have the answer. All questions and answers are automatically incorporated into the knowledge database.
Supporting Technologies: Portals, Collaboration Tools, and Learning Management Systems The major commercial knowledge management system vendors are integrating their content and document management capabilities with powerful portal and collaboration technologies. Enterprise knowledge portals provide access to external sources of information, such as news feeds and research, as well as to internal knowledge resources along with capabilities for e-mail, chat/instant messaging, discussion groups, and videoconferencing. Users can, for example, easily add a collection of documents obtained through a portal to a collaborative work space. The Gartner Group consulting firm uses the term Smart Enterprise Suites for this leading-edge knowledge management software. LEARNING MANAGEMENT SYSTEMS Companies need ways to keep track of and manage employee learning and to integrate it more fully into their knowledge management and other corporate systems. A learning management system (LMS) provides tools for the management, delivery, tracking, and assessment of various types of employee learning and training. A robust LMS integrates with systems from human resources, accounting, and sales so that the business impact of employee learning programs can be more easily identified and quantified. The first learning management systems primarily automated record keeping in instructor-led training. These systems have been enhanced to support multiple modes of learning, including CD-ROM, downloadable videos, Web-based classes, live instruction in classes or online, and group learning in online forums and chat sessions. The LMS consolidates mixed-media training, automates the selection and administration of courses, assembles and delivers learning content, and measures learning effectiveness. If a company had a customer relationship management (CRM) system that kept track of call-handling time, for instance, a sophisticated learning management system might be able to correlate performance data with training data to see whether training correlated with on-the-job performance. Recent versions of learning management systems with open architectures have capabilities for exporting their data to other systems. The Window on Organizations describes some of the benefits of learning management systems. Training for combat readiness and for job skills is an essential part of the U.S. Navy’s mission, and it must be conducted on a very large scale in many different settings and locations. Trainees have many different aptitudes, skills, and career paths to be managed. Trainees must be tested before and after they take courses. The Naval Education Training Command was able to implement a single learning management system that could handle all of these requirements. |