Extensible markup language and knowledge ... - Semantic Scholar

Introduction

Extensible markup language and knowledge management James R. Otto James H. Cook and Q.B. Chung

The authors James R. Otto is Assistant Professor, University of Baltimore, Baltimore, Maryland, USA. James H. Cook is Senior Science Advisor, IIT Research Institute, Lanham, Maryland, USA. Q.B. Chung is Assistant Professor, Villanova University, Villanova, Pennsylvania, USA. Keywords Knowledge management, Language, Documents, Internet Abstract Explores the use of extensible markup language (XML) to both store and enforce organizational data definitions, thus providing a synergetic framework for leveraging the potential of knowledge management (KM) tools. XML provides a flexible markup standard for representing data models. KM provides IT processes for capturing, maintaining, and using information. While the processes that comprise KM and the mechanisms that form XML differ greatly in concept, they both deal in a fundamental way with information. XML maintains the context of data (i.e. data model) which enables data to represent information. KM provides the framework for managing this information. Explores the vital role that XML can play to support an efficient corporate KM strategy. Electronic access The current issue and full text archive of this journal is available at http://www.emerald-library.com/ft

Journal of Knowledge Management Volume 5 . Number 3 . 2001 . pp. 278±284 # MCB University Press . ISSN 1367-3270

The prediction of Gordon Moore, a co-founder of Intel Corporation, that processor power would double every 18 months, known as Moore’s Law, has proved remarkably accurate as manifested by the exponential growth of the capabilities of microprocessor chips. Another interesting hypothesis suggested by Robert Metcalfe, founder of 3Com Corporation and designer of the Ethernet protocol for computer networks, provides a heuristic which speaks to the aggregate power of today’s networked economy. Metcalfe’s Law states that the usefulness or utility of a network is proportional to the square of the number of users (Downes et al., 2000). The impact of Metcalfe’s Law has been demonstrated by the explosive growth in the membership and utility of the Internet. Facilitated by the nonproprietary Internet standards, the continuing growth of the Internet fuels the unrestrained exchange of Web documents which enables the enormous number of the Internet users to both process and share information. It is easy to observe, however, that even given Metcalfe’s Law, the transfer of information encounters considerable friction due to differing representational standards. HyperText Markup Language (HTML), along with other Internet standards, provides the necessary nonproprietary standard that helped launch the explosive growth of the document sharing on the Web. However, while HTML is excellent at sharing Web documents, it is not ideal for exchanging information. Extensible Markup Language (XML), a published standard of the World Wide Web Consortium (Bray et al., 1998), promises to do for information what HTML has done for documents, thus reducing the friction and moving further toward realizing the full potential of Metcalfe’s Law. Figure 1 is a schematic representation of how information is shared between various business entities. It emphasizes the importance of an XML augmented Internet for information sharing between firms whether they are within the same industry or belong to different industries. Conceptually, information sharing between two firms within the same industry may be perceived as the bi-directional information flow shown as a dotted line arrow labeled (A). In actuality, it is most likely that such

278

Extensible markup language and knowledge management

Journal of Knowledge Management Volume 5 . Number 3 . 2001 . 278±284

James R. Otto, James H. Cook and Q.B. Chung

Figure 1 Information sharing within and between organizations

information sharing takes place via the Internet. That is, dotted line arrows (B1) and (B2) form a channel for information sharing between the two firms by way of the Internet, while the two solid line arrows, both labeled (B3), represent an intra-firm information sharing channel. In this scenario, establishing and enforcing standardized data definitions for intra-firm information sharing such as (B3) is fairly straightforward, whereas the inter-firm channel of (B1) and (B2) via the Internet or across industry boundaries such as (C) is burdened with inherent incompatibility and may require significant efforts to avoid the friction resulting from such incompatibility. Standard data definitions across a firm or industry are critical to increasing the number of users on the network (much like the introduction of standard railroad gauge revolutionized rail travel). Thus, standard data definitions may be a critical component for a future revolution in data exchange. Document type definitions (DTD) of XML may do for the exchange of data and context what HTML did for the exchange of documents. Simply put, XML can be used to reduce the transaction friction between information entities that may use different

data definitions and terms. By reducing such transaction friction, the capacity to share data elements and information across organizational boundaries may increase and result in larger, more efficient, and more valuable networks. The thrust of sharing information across organizational boundaries is that the parties involved will mutually benefit from increasing their competence in creating or capturing, retaining, disseminating, and exploiting knowledge, i.e. knowledge management (KM). KM is an emerging discipline with no agreed-on industry standard definition nor a framework in which to align different professionals (Ives et al., 1998). Regardless of definitions or common frameworks, Wiig (1997) suggests five knowledge-centered strategies, one of which can be adopted and pursued depending on the organizational objectives, namely: (1) KM as a business strategy; (2) KM as intellectual asset management; (3) KM to foster personal knowledge asset responsibility; (4) KM for knowledge creation; and (5) KM for knowledge transfer. Among the various themes that have emerged in the KM literature, a few imperatives stand

279

Extensible markup language and knowledge management

Journal of Knowledge Management Volume 5 . Number 3 . 2001 . 278±284

James R. Otto, James H. Cook and Q.B. Chung

out: e.g. formation of knowledge community, in which networks of people are considered the most critical element of successful KM implementation (Offsey, 1997; Martiny, 1998; Rowley, 2000), and focusing on innovation to realize the resultant benefits of KM (Brand, 1998; Pe´rez-Bustamante, 1999; Swan et al., 1999; Echeverri-Carroll, 1999). The need, plans, implementations and reports of results of KM is ubiquitous, ranging from the consulting industry (Martiny, 1998) to the pharmaceutical industry (Chase, 1997; Koretz and Lee, 1998) to law firms (Terrett, 1998). Both XML and KM are popular information technology (IT) buzzwords today since they have something in common. XML provides a flexible markup standard that can represent data models. KM represents IT processes for capturing, maintaining, or using knowledge. While the processes that comprise KM and the mechanisms that form XML differ greatly in concept, they both deal in a fundamental way with information. One of the primary responsibilities of data administration within a firm is to ensure data compatibility throughout the organization (McNurlin and Sprague, 1998). The data administration function must eliminate redundant and inconsistent data definitions because the data dictionary is a key tool for managing information resources across a firm (Charafas and Steinmann, 1993). Given consistently defined data, it becomes possible to combine and aggregate data from several sources and share it across the firm or compile it in a data warehouse. While it is possible to create organizational level data definitions, it is also the responsibility of the data administration function to enforce use of the data definitions. This paper explores the possibility of using XML functionality to both store and enforce organizational data definitions, and provides a synergetic framework for leveraging XML to realize the KM potential.

The role of content in KM In a discussion of the role of computer software, Armour (2000) identifies five media of knowledge, namely: (1) DNA; (2) brains; (3) hardware (tools);

(4) books; and (5) software. Knowledge in different media exhibits different characteristics. For example, knowledge in DNA is persistent but hard to update, while knowledge in the brain is volatile. In a perspective of history and development of civilization, these five media of knowledge form a cogent and smooth continuum. This view, however, fails to recognize one of the most notable and ubiquitous media of knowledge, the Web. It is hard to fit the Web into one of the five knowledge media identified by Armour (2000), and yet it carries vast amount of knowledge that we know of today. That is why managing Web content is high on the list of content management tool vendors’ key interests. By providing an infrastructure for creating and managing dynamic, evolving Web sites, Web content management tools can enable the organizations to perform their knowledgebased tasks (Aceto, 1999). According to the taxonomy of organizational knowledge proposed by Sackmann (1992), there are four categories of knowledge that is deeply embedded in organizational cultures; i.e. dictionary knowledge, directory knowledge, recipe knowledge and axiomatic knowledge. It is interesting to note that the Web technology (including the World Wide Web on the Internet, corporate intranets with proprietary contents, and extranets that establish connectivity between business partners) is rapidly progressing to serve as container of all of the four types of knowledge contents.

How XML’s DTD reduces information friction The primary concept that makes XML popular is its DTD, which standardizes the semantics and vocabulary of a domainspecific data model (St Laurent and Biggar, 1999). Specialized DTDs can be developed and agreed on by any organization such as a firm or by any group of organizations such as an industry or a supply chain, to facilitate data exchange between the parties involved. Table I shows a portion of a DTD that defines information relevant to a customer at an automobile dealer (source: www.dtd.com). Once a community agrees to publish data via

280

Extensible markup language and knowledge management

Journal of Knowledge Management Volume 5 . Number 3 . 2001 . 278±284

James R. Otto, James H. Cook and Q.B. Chung

Table I Sample DTD for an auto dealer customer

XML markup by referencing a common DTD, writing the software to create data management tools for the community is much easier. The services to access the community data set can now be provided by existing XML tools, exploiting code encapsulation and reuse. The effect is to make the context, i.e. data model, of a community data set available to any software application. XML also holds the promise of becoming a standardized mechanism for the exchange of data as well as documents. For example, XML may become a way for databases from different vendors to exchange information across the Internet. Furthermore, since XML, unlike HTML, separates data structure and content from presentation, a single XML source document can be displayed on a variety of platforms such as workstations, personal digital assistants, cell-phones, etc. (Gorman, 1998). Most recently, OnStar Inc. has revealed its XML-based voice delivery technology that can be used for weather forecast, news, sports scores and stock updates to wireless phones installed on 30 models of cars made by General Motors (Sliwa, 2000). Therefore, XML documents do not need to be modified as display devices and technologies change.

The role of context Corporations use KM tools to manage their knowledge assets. For instance, knowledge capture might be targeting the best practices

for customer management or raw data from laboratory experiments, utilizing Lotus Notes-based tools (Abramson, 1999). Knowledge maintenance and use might involve processes for storage and retrieval of these best practices or experimental data, utilizing tools such as Web-based data warehouses and knowledge repositories. Management of the tactical knowledge necessary for exploiting competitive intelligence (Vedder et al., 1999) might involve IT communications technologies. Tools used for such applications include groupware for marketing collaboration or Internet technologies such as spiders or bots to monitor and track external citations to competitor’s Web sites. XML can be used to maintain the context of information (i.e. data model) which differentiates it from data. KM provides the conceptual framework for managing this information. In this section, we will further explore these ideas and address the vital role that XML can play to support an efficient corporate KM strategy. Important to understanding the role of XML in KM is an appreciation of how data is related to information through context – which we define as meta-information that identifies the entities and their relationships within a data set. Here context is used to refer to the identity of, and relationships between, the entities of a data set. This data entity context is often referred to as a data model and can be represented by data dictionaries and entity-relationship (ER) diagrams.

281

Extensible markup language and knowledge management

Journal of Knowledge Management Volume 5 . Number 3 . 2001 . 278±284

James R. Otto, James H. Cook and Q.B. Chung

Information is essentially data with context so that data is associated with specific concepts. For example, 42 is an integer number (data). Associating that number with a concept such as the age of a specific person provides the context to turn that data into something relevant and useful (information). Without context, data becomes less useful. The tables in a database provide a contextual framework for data. Thus, when data is added to database tables, it becomes more useful because context is added (or maintained). Thus, KM can be defined as aggregating data, with context, for a specific purpose. Keeping the proper context associated with the data is a critical requirement of any KM process.

Why a KM infrastructure should utilize XML Since KM deals with the IT processes for working with corporate knowledge by aggregating information that addresses issues of importance to the corporation, the major functions to be considered here include capturing, maintaining, and using knowledge. The knowledge content varies from strategic information (such as best practices) to tactical marketing information (such as current assessments of a competitor’s market share), or even raw data from laboratory notebooks. In each case, wherever information is captured, processed, or stored, care must be taken to preserve or properly translate the context of the data that makes it information. Traditionally, KM stores and maintains context with data using databases. Unfortunately, this becomes unwieldy as the data stores must interface and interact with a wider variety of data types in addition to text (such as multimedia) and especially with more diverse information sources. This is especially true as information must be moved across entity boundaries, such as from Firm A to Firm B, or from Department 1 to Department 2 as shown in Figure 1. This is because these different entities may use different ways of defining context. Each networked party wishing to exchange information must understand the organization and schema of the other party’s data. This problem will only be exacerbated as more systems are interconnected for important applications such as electronic

commerce and supply chain management. XML can help solve this problem of preserving context across different database, corporate, and industry boundaries through the use of DTDs. A global XML DTD can provide the discipline and tools to create and enforce this organization-wide data context. Additionally, XML routines can be written to parse automatically a variety of knowledge sources to determine their compliance with the DTD standard.

Framework for implementing XML-based KM We believe that there are two major drivers for successfully incorporating XML into a corporate KM infrastructure: discipline and tools. Discipline is needed to develop and use a standard domain data model. Every participant must understand, accept, and properly use that part of the data model that impacts their activities. This requires the development of a consensus for the initial data model creation and careful monitoring to maintain that consensus. The design of the data model must anticipate future developments and the KM infrastructure must provide an efficient means for their incorporation. Since data models should be consistent across departmental and corporate boundaries, the development and management of the data model becomes a strategic level issue that needs to be supported at high levels within the organization. Tools are needed for users to easily interact with the data model. Fortunately, only the data analysts and the programmers who are directly involved in maintaining the DTD data model need to understand the XML and DTD syntax. Even the programmers creating the KM tools that access the data model do not need to know the XML syntax. They only need to know how to interface with the standard XML parsers that make that data model available to their software. For instance, a corporation may desire to capture all laboratory test results immediately at the source as XML markup to expedite mining the raw data. A domain data model component that addresses laboratory issues could provide this functionality. An example of an XML tool providing an interface to such an XML DTD would include organizational

282

Extensible markup language and knowledge management

Journal of Knowledge Management Volume 5 . Number 3 . 2001 . 278±284

James R. Otto, James H. Cook and Q.B. Chung

concepts and data items to provide a structure similar to the one shown in Figure 2. Tools to fit the laboratory raw test results to this data model immediately at the lab bench would involve augmenting existing automated data acquisition systems to make them XMLcognizant. Hopefully, the proliferation of XML-cognizant devices will follow the example of the proliferation of applications and devices that can interpret and present HTML. Such a proliferation of devices using a nonproprietary XML standard for sharing information (data and context) will do for information sharing what HTML has done for data presentation. Setting up the domain data model, creating new or modifying existing data collection systems, and enforcing their use will take discipline. The result, however, will be a tremendous improvement in the corporation’s ability to share, aggregate, and manage its diverse knowledge sources.

Summary and conclusions XML is often introduced in the literature as an extension of HTML. This is

understandable since both utilize similar markup notation. Those who understand HTML may be able to learn XML rather easily with this approach. However, this may put too much emphasis on presentation and mask the true value of XML. By facilitating common standards for context representation and thus reducing the friction of information flow, XML helps the Web become a more powerful information transfer medium. The ability of a DTD to provide access to a standard data model is fundamental to realizing this expanded power. XML, by using this mechanism to keep all corporate data connected with a standard domain data model, can play a vital role in the efficient sharing, aggregation, and management of a corporation’s or an industry’s knowledge. The key contribution of XML to a corporate KM strategy is to facilitate the process of maintaining this standard data model and keeping all data connected with that model. The most immediate impact will be to make the development and use of KM tools much easier. This will lead to the creation of more powerful tools to leverage a corporate KM strategy further.

Figure 2 An example of organizational concepts and data items as a basis of XML DTD

283

Extensible markup language and knowledge management

Journal of Knowledge Management Volume 5 . Number 3 . 2001 . 278±284

James R. Otto, James H. Cook and Q.B. Chung

References Abramson, G. (1999), ``On the KM Midway’’, CIO Enterprise Magazine, 15 May 1999. (http:// www.cio.com/archive/enterprise/ 051599_cons.html) Aceto, C. (1999), ``Managing Web content’’, Document World, Vol. 4 No. 5, pp. 18-23. Armour, P.G. (2000), ``The case for a new business model: is software a product or a medium?’’, Communications of the ACM, Vol. 43 No. 8, pp. 19-22. Brand, A. (1998), ``Knowledge management and innovation at 3M’’, Journal of Knowledge Management, Vol. 2 No. 1, pp. 17-22. Bray, T., Paoli, J. and Sperberg-McQueen , C.M. (1998), ``Extensible Markup Language (XML) 1.0’’ W3C Recommendation 10-Feb-98. Reference: REC-xml19980210. Charafas, D. and Steinmann, H. (1993), Solutions for Networked Databases, Academic Press, New York, NY. Chase, R.L. (1997), ``Knowledge management benchmarks’’, Journal of Knowledge Management, Vol. 1 No. 1, pp. 83-92. Downes, L., Mui, C. and Negroponte, N. (2000), Unleashing the Killer App: Digital Strategies for Market Dominance, Harvard Business School Press, Boston, MA. Echeverri-Carroll, E.L. (1999), ``Knowledge flows in innovation networks: a comparative analysis of Japanese and US high-technology firms’’, Journal of Knowledge Management, Vol. 3 No. 4, pp. 296-303. Gorman, T. (1998), Twenty Questions about XML, CNET, 10 March 1998. Available at: http:// www.builder.com/Authoring/Xml20/index.html Ives, W., Torrey, B. and Gordon, C. (1998), ``Knowledge management: an emerging discipline with a long history’’, Journal of Knowledge Management, Vol. 1 No. 4, pp. 269-74.

Koretz, S. and Lee, G. (1998), ``Knowledge management and drug development’’, Journal of Knowledge Management, Vol. 2 No. 2, pp. 53-8. McNurlin, B. and Sprague, R. (1998), Information Systems Management in Practice, Prentice-Hall, Upper Saddle River, NJ. Martiny, M. (1998), ``Knowledge management at HP consulting’’, Organizational Dynamics, Vol. 27 No. 2, pp. 71-7. Offsey, S. (1997), ``Knowledge management: linking people to knowledge from bottom line results’’, Journal of Knowledge Management, Vol. 1 No. 2, pp. 113-22. PeÂrez-Bustamante, G. (1999), ``Knowledge management in agile innovative organization’’, Journal of Knowledge Management, Vol. 3 No. 1, pp. 6-17. Rowley, J. (2000), ``From learning organization to knowledge entrepreneur ’’, Journal of Knowledge Management, Vol. 4 No. 1, pp. 7-15. Sackmann, S.A. (1992), ``Culture and subculture: an analysis of organizational knowledge’’, Administrative Science Quarterly, Vol. 37 No. 1, pp. 140-61. Sliwa, C. (2000), ``OnStar takes on voice XML for drive: GM unit prepares for mobile content delivery’’, Computerworld, 22 May. St Laurent, S. and Biggar, R. (1999), Inside XML DTDs, McGraw-Hill, New York, NY. Swan, J., Newell, S., Scarbrough, H. and Hislop, D. (1999), ``Knowledge management in innovation: networks and networking’’, Journal of Knowledge Management, Vol. 3 No. 4, pp. 262-75. Terrett, A. (1998), ``Knowledge management and the law firm’’, Journal of Knowledge Management, Vol. 2, No. 1, pp. 67-73. Vedder, R.G., Vanecek, M.T., Guynes, C.S. and Cappel, J.J. (1999), ``CEO and CIO perspectives on competitive intelligence’’, Communications of the ACM, Vol. 42 No. 8, pp. 108-16. Wiig, K.M. (1997), ``Knowledge management: an introduction and perspective’’, Journal of Knowledge Management, Vol. 1 No. 1, pp. 6-14.

284