Decision Support Systems 43 (2007) 1062 1079 www.elsevier.

com/locate/dss

A business process context for Knowledge Management

T.S. Raghu *, Ajay Vinze
Department of Information Systems, Center for Advancing Business through Information Technology (CABIT), W.P. Carey School of Business, Arizona State University, USA Available online 14 July 2005

Abstract Knowledge and management of it emphasize and expect interactions between aspects of business processes including workflow execution, information processing, decision making and motivational structure. As such production and consumption of knowledge occur within these aspects of business processes. Therefore, a business process context provides the justification and rationale for organizing Knowledge Management efforts that address knowledge storage and retrieval, knowledge sharing and knowledge synthesis. Exemplar projects are used to illustrate potential approaches and associated research challenges to addressing Knowledge Management efforts within a business process context. D 2005 Elsevier B.V. All rights reserved.
Keywords: Knowledge Management; Business processes; Coordination; Ontology; Knowledge sharing; Information systems

1. Introduction The challenge of managing knowledge in organizational context lies in effectively harnessing multiple knowledge sources into coherent business intelligence and embedding the intelligence into organizations memory. As the notion of Knowledge Management (KM) matures, it is increasingly clear that KM is not just about technology, and cannot be realized simply through information systems. Knowledge and management of it emphasize and expect collaboration between a wide spectrum of contributors that ranges
* Corresponding author. W.P. Carey School of Business; PO Box 874606; Tempe, AZ 85287-4606, USA. E-mail address: Raghu.Santanam@asu.edu (T.S. Raghu). 0167-9236/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.dss.2005.05.031

from people and processes to supportive technologies in an organization. Historically there has been a perceived disconnection between technologists and business managers as it relates to Information Technology (IT) solutions in business contexts. Technologists typically view IT from the perspective of capabilities, where the focus is on specific functionalities and interfaces afforded by the solution to the users. Business managers tend to evaluate IT solutions from strategic or business process enablement perspective [12]. In this paper, we suggest frontiers for KM efforts in the business domain. As has been recognized in a variety of disciplines, knowledge in the business context is a somewhat nebulous resource. To build a foundation for KM, efforts need to be directed for

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

1063

addressing issues related to knowledge storage and retrieval, knowledge sharing and knowledge synthesis. Using arguments grounded in existing theory and literature, we identify coordination mechanism for business processes as the knowledge context. Knowledge storage and retrieval provides a unique challenge to researchers as data management systems as a general rule use indices to store data and effectively the same indices are used to retrieve the data. As such orientations to storage and retrieval are closely tied together. For KM systems, storage approaches and retrieval mechanism are not necessarily congruent. A single level storage and retrieval mechanism cannot be applied in KM, necessitating an ontological orientation. The utility of this orientation is illustrated via a failure analysis/failure identification KM system developed in the context of a semi-conductor manufacturing organization. Knowledge sharing addresses the needs related to generation and collaborative aspects of knowledge. Knowledge artifactsdata and informationused in the sharing and generative processes are inherently
Table 1 KM research themes Time period 1960s1970s Driving forces Increased number of large organizations Transaction Processing Systems and Manufacturing Automation Salient themes

unstructured. Furthermore, these artifacts come from disparate sources causing the sharing process to be asymmetrical in orientation. We instantiate the complexity associated with knowledge sharing by presenting a real time knowledge sharing case in the context of bio-terrorism surveillance. Success in this setting is inherently predicated on effective and efficient knowledge sharing among sentinel data gathering sites, first responders and epidemiologists. The case illustrates the challenges that emerge from information asymmetries, unstructured data transfer and the opportunistic problem solving necessitated by discontinuous data, information and knowledge sharing. Before delving into our approach to creating a context for managing knowledge, it is instructed to briefly review the extant literature on KM. While the conceptual foundations of KM were laid out in the work of several management researchers in the early 1960s [18,35,46], more formal treatise of the concepts related to Knowledge began to emerge in the 1980s, especially as the resource based view of the firm began to gain currency [59]. The dominant theme in the KM literature in the

KM and information technology context Systems orientation to KM non-existent

Resource Based View of the firm [45] Knowledge Classification [46,35] Organizational Strategy [18] Organizational Learning Models [4] Bounded Rationality and Information Processing [39] Competitive Strategy Framework [48] Organizational design and Strategic Fit [40] Strategic Capability of the firm [49]

19801989

Globalization Shift toward service and Knowledge based organizations

1990 onwards

Total Quality Management Business process re-engineering Increased attention to Knowledge and intellectual capital management Emergence of Information Economy Tighter Inter-organizational relations in operations and strategy

Collaboration and communities of practice [24,44] Spiral model of knowledge creation; interaction between tacit and explicit dimensions [42] Industry practice and prescriptions for effective KM [15]

Expert Systems and Knowledge-based Systems in Research Labs (DENDRAL-1971; MYCIN-1975; HACKER-1975) Operational uses of Decision Support and Knowledge Based Systems [32] Computer Supported Cooperative Work (CSCW) [33] Group Support Systems [43] Internet based systems Data mining

Business Intelligence and Data warehouse technologies Enterprise Systems Document Management Systems Workflow Management Systems Intelligent and Mobile Agent Systems

1064

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

1990s reflected the increased emphasis on downsizing and responsiveness to customers, and therefore focused on reengineering and process-based view of the firm [27]. More recent literature in this area has explored several knowledge and KM related issues [1,3,9,16]. Themes in recent literature have explored mechanism design for optimal investments in knowledge [5], knowledge characteristics and organizational structure [8], knowledge creation and process change [13], knowledge reuse [36] and knowledge transformation [11]. A summary of the evolution of KM as a research area is provided in Table 1. Some of cited sources in the table provide a more detailed overview of issues related to organizational knowledge and KM. KM literature draws heavily from a related stream of research that has attempted to characterize and classify knowledge itself. This stream of research helps establish the rationale for varied approaches to managing knowledge due to the inherent complexities of knowledge artifacts. Polanyi [47] initiated the discussion of knowledge classification and introduced the notion of tacit and explicit knowledge types. By definition, tacit knowledge is acquired through experience and explicit knowledge can be acquired through articulation and codification [54]. Technology applications aimed at facilitating KM have been more successful in dealing with explicit knowledge. Attempts at facilitating tacit knowledge transfer and acquisition have focused on communication and networking technologies [6]. Research over the past few decades has identified Knowledge storage and retrieval, knowledge sharing (or transfer) and knowledge synthesis (or creations) as the three main facets of KM processes in organizations [11]. In the rest of the paper, we discuss the emerging facets of KM using these three facets as the main drivers of KM efforts in organizations. Our discussion argues for a clear application context for KM efforts and uses the coordination mechanism in business processes as this context.

been developed and studied in a variety of contexts as detailed in the previous section. For purposes of this paper, we restrict our focus to a business decision and process setting as we describe a knowledge context. Business knowledge is defined in the extant literature as a complex conglomeration of information, workflow, decision and collaborations and all the associated interactions. The inherent complexities and challenges for managing knowledge in a business context have been well documented. The KM literature in the business context, hence reported numerous successes though the extensibility of these successes to a wider context has revealed persistent problems [15]. Problems with KM and knowledge sharing are well documented and often result from lack of applicability of available knowledge. Often, technological solutions aimed at providing employees a means to exchange, publish and use knowledge related to their expertise fail for reasons completely unrelated to technology issues [23]. We argue that these problems arise when investments in KM processes and KM technologies are made without a specific knowledge context [58]. For this paper we define a knowledge context with an operational focus, where the knowledge unit and the KM efforts are intertwined and indistinguishable. Critical to this orientation is a definition of an operational context for knowledge and its application provided by the business process. 2.1. Business process as knowledge context Business processes are a collection of interdependent activities or tasks organized to achieve specific business goals. Business processes often cut across multiple functional organizations and hierarchies within and outside the organization and therefore require that the activities within the process be coordinated to achieve the business goals effectively. Researchers have attempted to gain a better understanding of business processes through the concepts of coordination framework [37,38,51]. Coordination is defined as arranging or organizing to achieve a desired or effective combination [55]. Coordination is achieved through mechanisms that are created to bind or organize the various aspects of a business process to meet the process objectives [50]. Extending the coordination framework developed by Raghu et al. [50], four key aspects form the

2. Knowledge context It is generally accepted that performance improvements from KM and associated technologies result when knowledge is actually applied [1]. Application of knowledge is to a large extent driven by its context which defines the intent of usage. KM efforts have

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

1065

knowledge context kernel as shown in Fig. 1: workflow execution, information processing, decision making and motivation structure. It is our assertion that all four of these components interact among each other and with KM processes in complex ways to result in an operational core of knowledge that should be addressed in any KM effort. 2.1.1. Workflow execution Activities and precedence relationships among the activities together form the workflow structure of a business process [29,56]. Workflow concerns usually revolve around issues of efficiency and flexibility. With the aim of increasing the workflow efficiency of a process, organizations typically focus on reducing handoffs, increasing concurrency or increasing automated tasks within the process [28]. On the other hand, organizations seeking to increase workflow flexibility focus on increasing the number of crosstrained workers and improved resource allocation mechanisms [10,34]. In either case, organizations face coordination and management problems when attempting solutions to efficiency and flexibility problems. For example, concurrency in workflow causes coordination problems that need to be solved through improved communication and information flow structures. Allocation of cross-trained workers requires careful consideration of the knowledge capabilities of the individual workers for optimal allocation. The utility of workflow in a KM context arises from several perspectives. Production oriented workflow processes such as customer support, claims

processing, call-center activities benefit from easy retrieval of past workflow instances of similar contexts. KM efforts directed at improving access and retrieval for such information can therefore improve workflow efficiency and consistency. Definition and dissemination of business rules that define the protocols of workflow instance execution improve the effectiveness of the workflow and perhaps, its efficiency. The definition of business rules requires effective knowledge capture from both within the organization as well as best practices from the industry. Given the dynamic and competitive nature of todays business, the capabilities of workflow structure within business processes take on added importance. Adaptation and knowledge synthesis within a workflow context occurs through complex interaction between KM processes and workflow structure. For example, to be able to learn from the past incidents in a technical support context, personnel should be motivated to document and detail the problem solving process in each incidence. Better documentation can potentially yield additional rules of interaction or opportunities for increasing concurrency within the workflow. However, documentation of the problem solving process interferes with workflow efficiency goals. Interestingly while the complementary nature of KM and workflow is undeniable, the contradictory goals of KM effectiveness and workflow efficiency present some interesting challenges to KM efforts. Issues related to efficiency and flexibility contribute to the operational core of knowledge in workflow

Knowledge Synthesis

Information processing

Motivation Structure

Operational Core of Knowledge

Workflow execution Decisionmaking

Knowledge Storage and Retrieval

Knowledge Sharing

Fig. 1. The business process context for KM.

1066

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

execution context. Workflow optimization efforts focus on achieving excellence in execution of routine tasks and thereby achieve efficiencies. Inherent in routine activities is the repeated interaction of employees over time with the business environment. Therefore, knowledge synthesis in a workflow context arises from systematic discovery of repeating patterns and creating an optimal set of activities to optimally address those patterns. Flexibility in workflow is achieved through a cross-trained workforce. However, differential levels of skills among employees can conflict with the goals of achieving efficiencies. Fortunately, Knowledge synthesis efforts that improve workflow efficiency can also enable flexibility through shared knowledge of patterns in workflow execution. When augmented with other means of systematic training and hiring policies, greater flexibility in workflow execution can be achieved. KM strategies and efforts in workflow execution contexts often begin with codification of explicit knowledge. Knowledge capture often occurs through structured database-oriented solutions that capture the explicit aspects of the problem solving process. Capturing the business rules and integration of the business rules within workflow execution engines provide the learning and adaptation context. The direct association of KM with workflow can result in higher order learning and adaptation through improved process monitoring and benchmarking practices. 2.1.2. Information processing The Information Processing (IP) paradigm from information theory [57] views organizations as being composed of units whose main function is to process information by exchanging messages among themselves [51]. A key objective of information processing activities is to respond to environmental uncertainty by processing and exchanging information. Additionally, the information processing paradigm [52,53] also takes into account the bounded rationality and bounded recall characteristics of humans. Since members of an organization are constrained by these cognitive limits to behavior, their goal is to conduct heuristic searches for satisfying decisions which meet a psychologically determined level of aspiration. Often business problems are based in a specific environment about which there is a lack of complete information. Therefore, specific interactions with the

environment have to be defined in order to explore and acquire additional information. Each of the information gathering action, however, has a cost associated with it. Additionally, in these problem solving situations where there is a computational cost, such cost needs also to be taken into account. The objectives of KM and IP run parallel in process design contexts, i.e., maximize the payoff by the right choice of action without excessive informational and computational costs [41]. Information processing functions in business processes are designed to minimize the computation , acquisition , and communication costs (all the three put together would be termed as the coordination costs of information processing). The design space includes information coordination, i.e., getting the right information to the right person at the right time to produce the right processed information, for example, in the processing of product design related information, budget planning, etc. In drawing parallels between IP and KM, information coordination is of utmost importance in processes that execute on a periodic basis, e.g., project planning, budgeting, strategic planning, product design, etc. Such processes often lack the rich patterns of learning that could be obtained in a workflow context. As such, knowledge storage and retrieval in information coordination contexts focus on storing documents, reports and computational aids (e.g., spreadsheet models) to reduce the cognitive burden on knowledge workers. The nature of stored artifacts facilitates knowledge reuse rather than knowledge synthesis. Knowledge synthesis in information processing contexts could be facilitated through knowledge sharing and business intelligence systems [2,19]. Computer Supported Collaborative Work (CSCW) systems and group systems are thus aimed at improving knowledge sharing through improved collaboration. At the outset, process goals and KM goals seem to be less contradictory in information processing contexts when compared to that of workflow execution. Business intelligence systems entirely built on top of transactional and enterprise oriented systems, and as well as external information sources are less susceptible to failure due to conflicts emerging from behavioral attitudes. This enables informational coordination through KM efforts to be less problematic. However, enabling and enhancing collaboration

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

1067

through KM systems is a much more challenging task. Lack of motivation among individuals and groups to participate in knowledge and information sharing activities has often been cited as among the primary causes for failure of CSCW and collaborative systems [17,21]. This can be attributed to the fact that in some cases there is a disincentive to participate in sharing activities altogether. The existence of enterprise systems and databases results in less of a need to invest directly in codification of knowledge for storage and retrieval purpose. However, it becomes essential to create at least rudimentary taxonomical knowledge structure when knowledge and data from a variety of sources have to be integrated and reconciled. Facilitating collaboration can benefit a great deal from both codification of knowledge and additional process related workflows to ensure that shared knowledge actually persists in organizational memory. 2.1.3. Decision making The impetus for creating explicit decision-making structures in business processes emanates primarily because it is not possible to centralize all the relevant information for decision making, expertise and control within a business process. The reasons for distribution of decision-making rights in a business process are due to bounded computational resources, and the limitations of time. The distributed decision-making structure in effect creates a form of organizational memory [60]. Decision-making components in a business process that aid in the achievement of process objectives and solutions include business procedures and business rules. Procedures govern the sequence of actions or heuristics that interpret and implement business rules. It is these procedures and rules that are intended to govern the actions of the members (or agents) in the day to day execution of the business process. While, workflow patterns can often embody business procedures, individual tasks also constitute enactment of business procedures, thus meriting separate consideration under decision-making structure. As will be discussed later, controls and motivational aspects of coordination mechanism are necessary to ensure that actions conform to the business rules and procedures. The decision-making structure and associated KM implications can be analyzed based on a dichotomous

classification of business rules and procedures as shown in Fig. 2. In Fig. 2, a decision-making structure that has formally defined business rules and where decision procedures follow the rules without room for alternative interpretations is termed rigid. Most bureaucratic processes can be characterized as being static and formal. For instance, the process of filling a prescription for a patient is a formal process, where the prescription has to be written by a practicing physician and has to fulfill pre-specified requirements, failing which the pharmacist will not fill the prescription. A decision-making structure which is evolving in procedure, and formal in the application of rules is considered as being oriented towards meaning and order. An example from the business domain includes Personnel policies, such as granting vacation time. While the formal rules are intended to ensure equity in the process, the procedures that implement the rules tend to evolve over time to allow some amount of flexibility in the interpretation of the rules. A decision-making structure comprising of static procedures, but informal rules are considered contextual and interactive . Tender processing is one example of such a process. The tender process follows a highly formal decisional procedure. However, the rules used for the decision choice are somewhat informal to accommodate better tenders, which may however not meet certain preset requirements. A decision-making structure which is evolving in procedure and informal in business rules is considered autonomous . This is the most flexible among the possible decision-making structure designs. Ideally, business rules and procedures for sales, marketing,
Contextual & Interactive Business Rules Rigid Formal Static Business Procedures
Fig. 2. Decision-making structure based on rules and procedures.

Informal

Autonomous

Meaning & Order Evolving

1068

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

purchase, and customer relations should fall under this category. Different business processes of an organization employ different decision-making structures to suit the needs of the objectives of the process being performed. Rigid decision-making structures tend to benefit the least from KM efforts. Any innovation or knowledge synthesis in the decision-making structure will have to be externally influenced through a change initiative. Decision-making structures emphasizing bmeaning and orderQ would benefit from KM efforts that drive innovations geared towards increasing the efficiency of procedural aspects of decision making. As with rigid decision-making structures, innovations and synthesis in business rules can only come through external influence. In essence, processes that emphasize formal business rules tend to depend on external forces for driving knowledge innovations in the process. We assert that technological solutions would have limited impact in such scenarios. Contextual and interactive decision-making structures depend on effective interpretation of informal business rules. KM efforts designed to better store and retrieve lessons learnt from past decision-making episodes would be beneficial in this regard. However, due to a static procedural orientation, knowledge within the context of the decision-making problem would be useful. There is very limited scope for learning from decisions made in contexts outside of the problem domain at hand. On the other hand, autonomous decision-making structures are the most amenable to knowledge sharing and storage and retrieval solutions. Such decision-making structures benefit from interactivity among decisionmakers both within and outside the process domain; they also benefit from retrieving knowledge related to solutions and procedures applied to similar decision problems from within and outside the problem domain. 2.1.4. Motivation structure A stream of research dealing with Agency relationships [30] recognizes that the decision makers (or agents) need not necessarily maximize the firm welfare under all conditions. In fact, this stream of research on agency theory stresses the importance of setting up appropriate incentive contracts to match the objectives of the individual decision makers with that

of the firm. The contracting relationships serve as a framework in which the conflicting goals of individuals are brought into equilibrium. However, all contractual models of human behavior have btransactionalQ and relationalQ components. The relational component emphasizes social exchange and interdependence; transactional component emphasizes the content of mutually agreed contracts [25]. To a large extent, organizational culture, trust and relationships are part of the relational aspects. Implicit psychological contracts govern agents perceptions of obligations and expectations of the organization [25]. The nature of the transactional and relational contracts that govern the behavior of the agents in the process leads to complex interactions among the other three aspects of the business process coordination mechanism. While impact of the transactional contracts on business processes is somewhat understood through agency theoretic models [30], the impact of transactional and relational contracts on KM efforts is very poorly understood. Unless the strong theoretical underpinnings of the interrelations are established, failures in KM efforts may continue to result due to a mismatch between the motivation structure and the aims of the KM efforts. 2.2. Facets of KM While the core of knowledge is defined through business process, the management of knowledge can be defined as a cyclical set of phases. While each of these phases can be a unique and self-contained facet of KM, it is the interactive nature of these orientations that accounts for the continuous evolution of knowledge and KM in organizations. We define the phases as the following: Storage and Retrieval, Knowledge Sharing and Knowledge Synthesis. 2.2.1. Knowledge storage and retrieval The relationship between storage and retrieval aspects of KM, and the business process is symbiotic. Effective knowledge storage practices enable storing of observations, consequences, and exceptions that occur in workflow execution and decision-making activities. Most effective storage practices tend to be non-intrusive and occur naturally as part of workflow and decision-making activities. However, when knowledge storage activities are external (and artifi-

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

1069

cial) to the business process context, effective control and motivational structures are necessary to ensure that the observations, consequences and exceptions are stored in the KM systems as per expectations. Furthermore, notions and concepts of storage developed from traditional data management practices are sometimes at odds with effective knowledge storage and retrieval practices. Traditional data management practices focus on redundancy reduction as an enhancement to storage practices. However, from a KM perspective, redundancy in storage often is a facilitator to better usage and synthesis of knowledge. When these perspectives differ and reflect on the different aspects of the problem scenario and the interpretations thereof, it actually enhances the inquiry and problem solving process rather than diminishing it. Therefore, while traditional data management practices emphasize accuracy through reduction in redundancy, KM practices may benefit from redundant information. Effective storage does not necessarily ensure knowledge synthesis or reuse during process execution through retrieval activities. To a great extent, this challenge in KM system usage stems from the nature of knowledge itself. In traditional data management systems, indices facilitate data and information retrieval. Storage and retrieval in data management systems mirror one another, i.e., data are stored and retrieved through the same set of indices. However, this is not necessarily the case when retrieval activities are performed in a knowledge context. Often, knowledge is stored using a set of indices (or a directory structure, in the case of documents) considered appropriate at the time of storing; but there is no assurance that these indices would be appropriate for the context within which knowledge may be retrieved. Thus, if the inefficiencies in knowledge retrieval activities are not addressed, the goals of knowledge reuse and synthesis may not be congruent with workflow execution and decision-making objectives. This incongruence would result in inappropriate use of KM systems. Data management practices typically emphasize accuracy in persistent transactional data. However, the notion of accuracy is difficult to define in a knowledge context. In fact, knowledge accuracy is always a function of user needs. The lack of definition of knowledge accuracy can lead to storage of inaccu-

rate knowledge. This is especially so, when motivational structure does not address knowledge storage or reward for contributions to knowledge synthesis through knowledge storage activities is minimal. Codified knowledge structures may address the accuracy problem to an extent by imposing rules for consistency within the knowledge structure. However, KM practices have to ensure that such knowledge structures can evolve over time to accommodate changes in the business process and technology environment. 2.2.2. Knowledge sharing Knowledge must be shared to be useful and applicable. When codified knowledge is stored, sharing is facilitated through defined access and security mechanisms and shared semantics of the stored knowledge. In cases where knowledge, however, is not systematically stored, it becomes necessary to create communication and collaborative mechanisms to enable knowledge sharing. This necessitates intense involvement of knowledge workers in communities of practice. Codification of the knowledge domain through at least simple taxonomical structures enables the search for expertise within the social network. Knowledge sharing in this context would depend to a great extent on the motivational aspects within the process context and the cultural setting within which the process operates. It is useful once again to draw parallels to traditional data management practices. Sharing or integrating data in traditional data management systems increases complexity (as in distributed database systems) through requirements such as simultaneous updates and locking. In knowledge sharing systems, increased size usually increases the utility of the knowledge shared through externalities. With increased size of the repository, structuring of data becomes a necessity for data sharing in data management systems through relational, network or hierarchical approaches. Knowledge sharing, however, can thrive even in the absence of structure so long as a context for the shared knowledge exists. Structured data and informational focus in data sharing facilitates queries on the relations that exist within data and summarization. Informational focus in knowledge sharing contexts emphasizes inquiry where search is opportunistic. Querying within a data sharing context has the notion of an accurate end-state (e.g., factual

1070

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

information). Inquiry in a knowledge sharing context may not necessarily involve an end-state. When an end-state exists, it is usually in the form of a solution to an unstructured problem with no verifiable (and perhaps, single) true end-state. 2.2.3. Knowledge synthesis The intent of storage and retrieval systems, and knowledge sharing systems is to enable and enhance knowledge synthesis capabilities of the organization. Yet, complex interactions between the KM facets of storage and retrieval, and sharing with process coordination mechanisms may not always create conducive environments for knowledge synthesis. While workflow and information processing aspects of business processes may create environments for effective storage and sharing practices, the decision making and motivation structure would have a critical impact on knowledge reuse and knowledge synthesis. Rigid decision-making structures clearly impede any innovative practices in business processes, thereby effectively nullifying the benefits of well-designed KM systems. Such processes often have to depend on innovations from outside the process boundaries. We, however, contend that KM systems have minimal impact on knowledge synthesis in processes characterized by rigid decision-making structures. Processes that emphasize contextual and interactive decision making clearly are conducive to knowledge reuse. Such processes can use the existing knowledge stores and knowledge sharing facilities to enhance the applicability of codified knowledge and increase decision-making efficiencies. Innovations in such contexts come through novel modifications to the rules of engagement and execution to maximize process objectives. Procedural adherence can however severely limit the innovation capabilities of the process. In a similar vein, decision-making structures emphasizing meaning and order to the process encourage knowledge synthesis through evolutions in decision procedures, however, are hampered by adherence to formally defined business rules. Processes emphasizing autonomous decision-making structures present the most conducive process environment for knowledge synthesis as they are inherently designed to allow innovations in procedures and business rules. KM efforts in such processes would have the most impact in enabling

knowledge synthesis through effective sharing, and storage and retrieval practices. We use two different cases to illustrate the challenges and the attempted solutions towards KM. The cases illustrate that organizations can tailor their approach to KM by increasing the relative importance of any of the three facets of KM, i.e., knowledge storage and retrieval, knowledge sharing, and knowledge synthesis to best address specific needs. The first case details efforts in a large manufacturing organization. KM attention for this case is focused on failure analysis and failure identification (FA/FI). Knowledge storage and retrieval issues are covered as we highlight the KM challenges and approach. In the second case we detail knowledge sharing challenges in a large public sector organization focused on health care services.

3. Lessons from the field 3.1. The storage and retrieval conundrum The organization (here on referred to as Mfg. Inc.) is a successful and entrepreneurial organization that is among the foremost leaders in its industry. KM is viewed as a critical activity since manufacturing process efficiency and innovation drive success in this industry. Tightly integrated into the manufacturing process is a highly responsive quality assurance program augmented by sophisticated lab analysis and resolution capabilities and staff. These labs respond within hours or days to any and all problems and evaluation requests from manufacturing and testing processes. Facilities for R&D (that drives design and manufacturing process innovation), testing and production are widely distributed across the globe. The time frame from development and validation of specific changes through production process to actual production may span several years. Mfg. Inc. encourages a highly entrepreneurial spirit within its process design teams. As product and process innovations are tested and sent to analysis labs, various reports are developed. However, these reports are loosely standardized. Report formats are individually determined and generally focused on a specific individuals or teams purpose. The challenge to Mfg.

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

1071

Inc. is in allowing those in other labs, particularly those labs who will continue production at a much later time, to retrieve relevant failure analysis and failure identification (FA/FI) reports eliminating redundant analysis, thereby optimizing productivity. Thus, the focus of the KM effort began with enhancing the storage/retrieval practices of the FA/FI process. The goal of the effort was to make better use of existing knowledge base on failure causes as well as analysis tasks to reduce throughput times in the analysis labs. This, in essence was the process context for the KM effort. It was evident that workflow execution and information processing aspects of the process were at the core of the issues affecting the KM effort. From the time of the failure to the final determination of root cause, personnel of various units and analytical roles would have recorded various reports and documentation. Each lab unit, however, has its own vocabulary, distinct from other manufacturing process and quality assurance groups. This somewhat autonomous nature among groups led to homonym or synonym discrepancies. For example, one group would use the term defect to indicate a condition influencing the failure. Another lab would only use the term defect to mean an anomalous condition specifically relating to an electrical stress failure. In another example, one lab used the term root cause to indicate the condition influencing the failure, such as a structural defect; whereas another group used the same term to indicate the source of the anomaly, such as material impurity. We utilize ontology as an enabler for creating the multiple foci. When an individual considers the domain ontology, the target knowledge element is framed by a series of presumed concepts and terminology interrelated to provide meaning thus facilitating effective retrieval of stored knowledge. Ontology is the basic structure or armature around which a knowledge base can be built [31]. A simple ontology may include a hierarchy of concepts bounded by subsumption relationships. More complex ontologies include axioms to increase the complexity of relationships, concepts, and constraints desired to fully bind the intended interpretation. The ontology consists of a conceptual model, a thesaurus, and a set of expanded attributes and axioms. Its concern is for the appropriate representation of content, which may

later be augmented using specification languages such as UML, RDF, BNF, or formal logic [20]. For Mfg. Inc., we built the control vocabulary from archival workflow data and reports, as well as from semi-structured interviews. The focus for this effort was the analysis lab where images and textual files formed reports of structural and chemical analysis during test and production runs. The analysis reports numbered in the thousands, so to refine the analysis to a workable set, a recently validated process was chosen. We addressed problem analysis from the perspective of a specific analysis lab with regard to the chosen process, and set out to develop an ontology to represent this perspective, with the objective of operationalizing the ontology in the FA/FI workflow to facilitate the seamless integration of storage and retrieval activities. In constructing ontology, one of the main challenges becomes bwhere to start?Q In identifying the main elements for the ontology, we looked at the workflow of reports first. These elements revealed entities, relationships, and attributes to be included within the ontology. Entities and relationships were extracted to develop a graphical, conceptual model. To facilitate understanding, the conceptual model was reduced to a series of sub-models organized around the primary elements extracted from the reports. A sample model is shown in Fig. 3. These sub-models depicted subsumptive and inter-element relationships. The graphical nature of the model benefited a wider range of users in that it is easier and quicker to understand than either textual or computational representation. Next the task of seamless integration into the process workflow was addressed by first developing a formal interview structure, initiating formal interviews, and incorporating new knowledge elements within the ontology model [20]. Revisions were reviewed with participants repeatedly. The model was evaluated as sufficiently robust to provide immediate value and opportunity for subsumption within a workflow tool under development. For initial evaluation of the utility of the new storage/retrieval capability, a partial ontology was encoded within a retrieval tool and evaluated for utility by comparing it with a similar environment sans the ontology. Most participating lab analysts preferred the tool in which more of the ontology was embedded over the simple free text search tool,

1072

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

Root Cause Ontology

failure

Caused by Results from

Root cause

Confirmed by

Working group or Task team

Originates from Anomaly Design cause

Originates from

Manufacturing cause

Unaccounted cause

Is_a Engineering design

Is_a Manufacturing process design

Found in

documents

Wafer

report

Fig. 3. A sample ontology extract from FA/FI process.

because of the improved retrieval capability. From initial impressions and qualitative feedback, the process of knowledge discovery was noticeably improved by placing the ontological structure on the information. However, to fully exploit the capability of the improved knowledge storage/retrieval facility, it should be closely integrated with process workflow. In fact, in Mfg. Inc., the refinement of the workflow system development was proceeding in parallel with the storage/retrieval enhancement effort. Lack of a uniform vocabulary impeded efforts to capture the meta-data on failure events and reports. The creation of the knowledge ontology was to enable effective sharing of process knowledge between the different functional units actively involved in the FA/

FI process. The knowledge ontology was also intended to facilitate redesign of workflow systems to make it easier to capture context-specific information on failure events. Thus the knowledge ontology about failure created solutions for effectively solving the storage and retrieval conundrum thereby paving the way for knowledge synthesis within the manufacturing process. While hard evidence of successes from this effort is hard to quantify at this stage, anecdotal evidence indicates better use of the knowledge store in failure resolution process and better information processing capability at the individual analyst levels. Mfg. Inc. provided a scenario where defining the business process context for KM efforts was quite straight forward. The FA/FI process constituted the

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

1073

context on which the KM effort was directed at. It enabled a clear business value justification for the KM effort and enabled quick buy-in from key process owners. Interactions between workflow execution and Information processing aspects presented challenges as well as opportunities to improve knowledge synthesis and reuse within the FA/FI process. Workflow execution in the FA/FI process caused some details of the failure context or event to be either misrepresented or understated. This often caused problems later in the process where initial information gathered about the failure event had to be either discarded or modified. This often reflected in a mismatch between eventrelated data stored in traditional database format and the information contained in the technical/analytical report that documented the failure analysis. Over the course of time, it became difficult for analysts to relate reports to specific events and vice versa. Thus, workflow execution structure impacted information storage. Partly as a result of this, the search process for past failure events related to a current failure event was cumbersome. Therefore, knowledge reuse through past lessons, despite a comprehensively archived storage and retrieval system, was challenging. Processing information related to failure event often occurred through informal channels which would later make it difficult to document all aspects of the problem resolution process. The solution envisaged for the FA/FI process included an overhaul of the workflow execution structure to provide better means to store failure related data in traditional database formats as well as FA/FI reports. FA/FI Ontology helped create a structure and template for documenting the FA/FI process related to specific events. Past failure reports were re-formatted in accordance with the ontological structure discovered during the KM effort (this was accomplished using a commercially available tool). The newly imposed structure enabled better information retrieval by means of both directed and keyword oriented searches. Information retrieval also benefited from additional linkages created between documents within the context of specific searches. 3.1.1. KM implications Success with knowledge storage and retrieval effort is often predicated on characterizing the knowl-

edge unit. A common orientation to characterization of knowledge unit adopts the notions of control vocabulary and taxonomy. Knowledge includes considerations for process (e.g., business process), assumptions, and justifications which are only partially subsumed within vocabulary and taxonomy. Additionally, unlike data and information, knowledge retrieval requires multiple foci to accommodate diverse yet legitimate perspectives of the desired knowledge. In many organizations failure analysis and failure identification (FA/FI) knowledge, or what is hoped to be knowledge about failure, is stored within structured database systems, within documents, within email messages, or as randomly saved files. The rationale for storing this knowledge is to be able to synthesize additional knowledge from the failure incidents and thereby improve process performance. The FA/FI perspective identifies failure not merely as a discrete event requiring a discrete, corrective response, but a context and motivation from which new forms of knowledge are created opportunistically. Effective knowledge synthesis in FA/FI context requires that the knowledge uncovered from failure detection be represented within a context that increases knowledge sharing within and across business processes. The FA/FI process ontology project at Mfg. Inc. generates quite a few challenging research issues from the perspective of creating KM processes focused on effective knowledge storage and retrieval efforts. FA/ FI processes and concerns resonate across multiple domains. For instance, in the IT services domain, complexities of enterprise IT systems preclude IT services of the possibility of fault-free operations. As such, error detection and resolution must occur within minutes or hours for critical problems affecting customers and supply chains. Failure, unlike in the manufacturing context, is not detected but traced from an anomalous result or incorrect code execution termination. Failure in a public health context is measured in the number of deaths attributed to the epidemic or outbreak. Despite the apparent differences, many domain concepts of failure can be given domain independent contexts. The cross-domain constructs enable one to develop a unified model of failure events and knowledge units related to FA/FI from a domain independent perspective.

1074

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

Emerging research efforts in Knowledge storage and retrieval contexts will be dominated by two key concerns: (1) identification of KM needs in commonly occurring business processes across domains, and (2) development of cross-domain knowledge ontologies within the context of specific but commonly occurring business processes across domains. Creation of crossdomain knowledge ontologies will require us to answer challenging research problems including a definition of bKnowledge UnitQ when the intent is to develop a bmetalanguageQ of the core elements of the business process knowledge to create a cross-domain knowledge ontology representing domain-independent knowledge concepts; definition of measures to evaluate cross-domain knowledge ontology. Cross-domain ontologies should be easily transportable (from one domain to another); flexible (ontology should be extensible); and maintainable (open to modifications in domain or cross-domain process knowledge); evaluation of the domain-independent knowledge ontology to the domain-specific specializations with respect to transportability, flexibility, maintainability, and elicitation value. 3.2. Knowledge sharing challenges Following the anthrax episodes of 2001, public health departments, and their preparedness and ability to respond to biological and chemical attacks have become the focus of attention across the nation. Public health departments are at the core of federal and state efforts to enhance the public health infrastructure to ensure better preparedness and response to deal not only with a potential attack of biological or chemical terrorism, but with any other large-scale health emergency. Federal response to the bio-terrorism threat resulted in new legislation and funding avenues to strengthen the public health infrastructure. As part of the federal response, the Centers for Disease Control (CDC) was mandated with helping states and local agencies achieve the required levels of preparedness for bio-terrorism surveillance. Knowledge sharing especially as relates the operational core of knowledge is at the center of the public health challenge of our times. Communication and connectivity form the cornerstones of knowledge sharing and can be described as the enablement of flow of the bright knowledge to the right person at the right

timeQ [26] that allow organizations and individuals to explore the inherent challenges that arise from the operational core of knowledge and to design and approach in concrete ways strategies for a response. The public health system comprises a vast network of local, state, and federal organizations. These entities are further divided along functional lines such as health departments, laboratories, disease programs, clinics, and other operational divisions. A coordinated response to any large-scale emergency requires the various units of the public health system to work collaboratively. Knowledge generation and usage is a rapidly evolving process, especially when a major emergency is in progress. Management of knowledge in public health as relates emergency preparedness and response is dependent in significant ways on the synergies spawned through knowledge sharing. The knowledge sharing in this context is typically across organizational boundaries, for example, workflow process execution is achieved through medical labs, hospitals and first responders working in close partnership. In non-emergency situations, these three stakeholders report to and are part of very distinct organizations. Information processing likewise needs to be coordinated between the first responders, emergency management personnel and the policy makers in a process that is very loosely defined and highly distributed. The distributed and decentralized nature of workflow and informational processes brings significant strain on decision making and motivation structures and thus increases the importance of knowledge sharing in this context. Concerns about the danger of bio-terrorism are high because of the perceived ease of transporting and deploying biological and chemical agents, the numerous avenues through which these agents can be released into the public domain, and once released, the possibility of rapid disbursement within the community. Thus, improving preparedness for, and response to public health emergencies such as bioterrorism is no longer a matter of choice; it has become an imperative for the public health system in this country. Strengthening the public health systems ability to deal with a bio-terrorism emergency will also result in better preparedness to deal with other emergencies such as chemical and nuclear terrorism, naturally occurring infectious diseases, and other mass casualty events.

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

1075

Knowledge sharing was also defined as a key success factor in tackling bio-terrorism surveillance. CDC defined the process of knowledge sharing as focus areas (see Fig. 4) in which critical capacity must be achieved for developing effective epidemiological and bio-terrorism surveillance systems at all levels of public health administration. Working with public health officials and organizations at the state and county level in a southwestern state in the US an approach was devised to address the issues for emergency preparedness and response that were raised by CDC. The framework for the approach is shown in Fig. 4. The approach devised and detailed aims at bringing efficiencies through a KM orientation. Knowledge sharing in bio-terrorism context also highlights an interesting constraint in decision-making structure. Surveillance processes in public health agencies have historically been characterized by brigidQ decision-making structures. Formal rules define the interactions between agencies, and while static decision procedures are used to act upon the shared information. Terrorism threats have necessi-

tated creation of further enacting of new rules for interactions between agencies. Increasingly, however, agencies have evolved their decision procedures to create bmeaning and orderQ in their decision-making structure. Knowledge sharing and communication are central elements of these evolving decision-making structures. These change efforts are complemented through development of a comprehensive information strategy that will bolster the public health and medical response to bio-terrorism and help integrate various components and different detection and response agencies into a single system. It is important to note that in this effort, the focus is not only on modern and traditional technologies, but also on human organizations and communication protocols that enable seamless information sharing to take place. Newly detailed complex interactions among numerous local, state, and federal agencies in the public health system necessitate a robust information technology (IT) infrastructure. The infrastructure needs to be deployed with well-defined information sharing methodologies to ensure a coordinated response to bio-terrorism and other public health issues. Public

Organizing Knowledge Sharing to Address Bio-Terrorism Challenges

Communication Connectivity

B
Integrated/Interoperable database for disease surveillance system Shared systems across different databases Data Mining models

Workflow Process and Information Processing

Effective risk communication Timely information dissemination Training and online-learning

E
Communication systems redundancy Critical data and information security Cross-platform connectivity Vocabulary and standards specifications

Focus Areas for Bio-Terrorism - CDC

A) Preparedness planning and readiness assessment B) Surveillance and Epidemiology capacity C) Laboratory capacity - biologic agents D) Laboratory capacity - chemical agents E) Health alert network/communications and information technology F) Communicating health risks and health information dissemination G) Education and training

A
Strategic Orientation

Fig. 4. Knowledge sharing for bio-terrorism surveillance.

1076

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

health events do not always stand out as obvious outbreaksin most cases an event profile is generated via active data and information sharing. A series of unrelated stomach illness incidences in a large metropolitan area can be evaluated as injection of a contaminant in the water supply. For timely diagnosis of the events, labs across the city, emergency room physicians and public health personnel have to collaborate in real time. To facilitate this process of knowledge sharing, the state of Arizona, in collaboration with the counties, cities and related first responders and labs are creating a system, Medsis. Medsis is intended to facilitate knowledge sharing to reduce time to diagnose a bio-terrorism or other medical events. CABIT efforts played a complementary role in this context. The CABIT report parallels the report on IT strategy for bio-terrorism by the United States General Accounting Office identified 70 planned and operational information technology systems within six key federal agencies related to public health emergencies [7]. These IT systems included detection, surveillance, communications systems and supporting technologies. Recognizing the critical role of information technology in bio-terrorism preparedness and response, the CDC has defined a public health information architecture that includes detailed specifications on data, communications, and security standards. Through knowledge sharing public health can capture, store, manage, analyze, and disseminate relevant health related information in emergency situations. 3.2.1. Implications and emerging frontiers Knowledge sharing, especially, in inter-organizational mission-critical decision-making scenarios, is critical to timely and effective resolution to decisional problems. Our approach to knowledge sharing in the bio-terrorism context highlights technology and process solutions that may aid in the process of knowledge sharing across organizations. The problems of motivational structures in the business processes exacerbate the impediments to knowledge sharing and knowledge synthesis in this context. Emerging KM solutions should develop critical decision-making mechanisms necessary to reduce cognitive dissonance among decision-makers. This is especially relevant in scenarios where decisionmakers are geographically and organizationally dis-

persed and are concerned with rapidly evolving situations. Such decision problems are often ill-structured and intractable due to the multitude number of events, evidences and facts that require careful consideration. Such scenarios are not necessarily limited to bioterrorism surveillance (although the potential disaster magnitude of the domain makes it a high impact domain) and occur even in business environments (such as security breaches in IT systems, disruptions in supply chains, etc.). KM efforts in such mission-critical decision scenarios should provide support in the decision-making processes that are involved in preparedness and prevention, event recognition, early and sustained response, and recovery (in the event of failure) [22]. Challenges to effective preparation, surveillance and response to these events include lack of shared semantics among various agencies and decision-makers, poor communication protocols, and distributed islands of domain knowledge, business intelligence information, and subject-matter experts. Procedural impediments to knowledge sharing can be removed through mandated workflow and structural changes to the process. For example, as a response to bio-terrorism threats, to address poor understanding of semantics and communication protocols, coordinated changes at the federal agencies are being effected through current and planned efforts. However, providing technology support for effectively integrating, synthesizing and managing the distributed islands of knowledge, information, media and subject matter experts is a KM challenge that will have to be effectively addressed by researchers in the immediate future. In part this challenge can be addressed through effective storage and retrieval mechanisms, more specifically, through a unified and federated framework of knowledge ontology representation that addresses and solves the problems of semantic interoperability among heterogeneous knowledge sources. Building communities of decision-makers that interact often is a necessity to create effective knowledge sharing processes. While communities are viewed as an important aspect of knowledge sharing initiatives, effective and vibrant communities are a rarity. The challenges faced by communities seem to be due to limitations of both technology and social context. The collaboration technologies should reflect the social protocols that underlie group communica-

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079

1077

tion in terms of strategies and policies for decision making.

4. Conclusions and implications While historically, technological innovations have tended to lag and later support business innovations (as in the case of six-sigma initiatives and just-in-time replenishments), more recent innovations in technologies tend to spur and generate innovative business practices (e.g., e-business practices and Radio Frequency ID tags) and thereby reverse the trend in instigating innovation through knowledge. Traditional view of knowledge as data and information fails to incorporate process and associated assumptions thus causing loss of context for the knowledge that is stored (or retrieved), and shared. Using the business process as a knowledge context provides a richer and more complex phenomenon to study. The business process context also provides an assessment framework that makes it easier to evaluate the impact of KM efforts in improving business process performance. KM itself is embedded in an iterative process that fluctuates between storage and retrieval, and knowledge sharing; with the ultimate aim of knowledge reuse and knowledge synthesis. The iterative view of KM enables the tailoring of KM efforts to the process needs. For instance, in Mfg. Inc. project, we initiated KM efforts from the perspective of improving knowledge storage and retrieval. In the bio-terrorism context, the focus was on improving knowledge sharing among multiple organizations. Yet, in both projects, the ultimate aim was to synthesize new knowledge from the two different perspectives. From a storage and retrieval perspective, interactions between workflow execution and information processing aspects present challenges as well as opportunities to improve knowledge synthesis and reuse. When efficiency concerns in workflow execution cause details of events and contexts to be either misrepresented or understated it impedes knowledge synthesis. When workflow design does not facilitate creation of appropriate relational structures between workflow data and documents, it becomes difficult to reuse the knowledge generated in the process at a later time. Thus, poorly designed

workflow execution structure can impact on information storage and make search and retrieval activities cumbersome. A feasible solution to this problem can be achieved through a well-designed ontological structure for the process context. The ontological structure helps to streamline workflow execution while simultaneously facilitating effective storage and retrieval of contextual information and knowledge pertaining to significant events in the process. In public health, decision-making structures are rigid since they are based on formal rules of engagement and static procedures. Typically these are delivered in the form of mandates or rules for regulatory agencies and external stakeholders, like the CDC and others. The ground level realities however require a much higher level of flexibility through knowledge sharing. As such decision-making structures that emphasize bmeaning and orderQ through evolving business procedures are essential. For situations typically found in public health, policy, rules and regulations significantly frame the knowledge unit or artifact. It is these rules and regulations that are supposed to facilitate knowledge sharing as part of standard operations within public health. In rapidly changing and emerging environments presented by bio-terrorism reaction and response, the very mechanisms that are supposed to facilitate the management and sharing of knowledge severely hamper the decisional situation. It is in situations like these that technology can be a key enabler. Using the knowledge sharing perspective, especially in multi-party decision-making scenarios, it is crucial to address the motivational and decision-making orientations. To be able to effectively synthesize knowledge through sharing, perceptual differences in interpretation of events and evidences in the decision scenarios need to be actively addressed. Finally, as KM and associated issues evolve we need to recognize and respond to the need for a cross-domain orientation to enhance value contribution for organizations by KM. When a process oriented view is taken towards KM, moving from one domain to another seems to be more plausible than when a purely functional orientation is employed. Across multiple domains, as an example, failure analysis and failure identification processes should view failure events as not merely a discrete event

1078

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079 [16] T.H. Davenport, et al., Data to knowledge to results: building an analytic capability, California Management Review 43 (2) (2001). [17] A.R. Dennis, et al., Information technology to support electronic meetings, MIS Quarterly 12 (4) (1988 (Dec)). [18] P. Drucker, The Age of Discontinuity: Guidelines to Our Changing Society, Harper and Row, New York, 1969. [19] P.F. Drucker, Knowledge-worker productivity: the biggest challenge, California Management Review 41 (2) (1999). [20] T. Edgington, Beom-Jin Choi, K. Henson, T.S. Raghu, A. Vinze, Ontology-Enabled Knowledge Management: Adopting Ontology to Facilitate Knowledge Sharing, Communications of the ACM 47 (11) (2004). [21] M. El-Shinnawy, A.S. Vinze, Polarization and persuasive argumentation: a study of decision making in group settings, MIS Quarterly 22 (2) (1998 (Jun)). [22] Gilmore-Commission, The Advisory Panel to Assess Domestic Response Capabilities for Terrorism Involving Weapons of Mass Destruction, Fifth and Final Annual Report, Rand Corporation, Arlington, VA, 2003. [23] D. Gilmour, How to fix knowledge management, Harvard Business Review 81 (10) (2003 (Oct)). [24] P.S. Goodman, E.D. Darr, Computer-aided systems and communities: mechanisms for organizational learning in distributed environments, MIS Quarterly 22 (4) (1998). [25] D.E. Guest, Is the psychological contract worth taking seriously? Journal of Organizational Behavior 19 (1998). [26] C. Gwin, Knowledge Sharing: Innovations and Remaining Challenges, IBRD/The World Bank, 2003. [27] M. Hammer, J. Champy, Reengineering the Corporation: A Manifesto for Business Revolution, 1st ed., Harper Business, New York, 1993. [28] M. Hammer, S. Stanton, How process enterprises really work, Harvard Business Review 77 (6) (1999 (Nov/Dec)). [29] G.P. Huber, R.R. McDaniel, The decision making paradigm of organizational design, Management Science 32 (5) (1986). [30] M.C. Jensen, W.H. Meckling, Theory of the firm: managerial behavior, agency costs and ownership structure, Journal of Financial Economics 3 (4) (1976). [31] H.M. Kim, M.S. Fox, Towards a data model for quality management web services: an ontology of measurement for enterprise modeling, in: P.A. Banks, et al., (Eds.), Caise, LNCS, vol. 2348, 2002. [32] B. Konsynski, R.H. Sprague Jr., Future research directions in model management, Decision Support Systems 2 (1) (1986 (Mar)). [33] K.L. Kraemer, J.L. King, Computer-based systems for cooperative work and group decision-making, Computing Surveys 20 (2) (1988). [34] A. Kumar, W.M.P.V.D. Aalst, E.M.W. Verbeek, Dynamic work distribution in workflow management systems: how to balance quality and performance, Journal of Management Information Systems 18 (3) (2001/2002 (Winter)). [35] F. Machlup, The Production and Distribution of Knowledge in the United States, Princeton University Press, Princeton, NJ, 1962.

requiring a discrete, corrective response, but a context and motivation from which new forms of knowledge can be opportunistically created. As organizations become more global and/or virtual, a unifying, semantically developed structure to represent knowledge becomes increasingly imperative. KM is an emerging frontier for the IS discipline. The promise of KM for the IS discipline is to bring harmony to the technology capability and possibilities.

References
[1] M. Alavi, D.E. Leidner, Review: knowledge management and knowledge management systems: conceptual foundations and research issues, MIS Quarterly 25 (1) (2001). [2] L.M. Applegate, J.I. Cash, Q.F. Mills, Information technology and tomorrows manager, Harvard Business Review 66 (6) (1988). [3] L. Argote, P. Ingram, Knowledge transfer: a basis for competitive advantage in firms, Organizational Behavior and Human Decision Processes 82 (1) (2000). [4] C. Argyris, Single-loop and double-loop models in research on decision-making, Administrative Science Quarterly 21 (3) (1976 (Sept)). [5] S. Ba, J. Stallaert, A.B. Whinston, Optimal investment in knowledge within a firm using a market mechanism, Management Science 47 (9) (2001). [6] M. Bieber, et al., Toward virtual community knowledge evolution, Journal of Management Information Systems 18 (4) (2002). [7] Bioterrorism: Information Technology Strategy Could Strengthen Federal Agencies Abilities to Respond to Public Health Emergencies, US General Accounting Office, Washington, DC, 2003. [8] J. Birkinshaw, R. Nobel, J. Ridderstrale, Knowledge as a contingency variable: do the characteristics of knowledge predict organization structure? Organization Science 13 (3) (2002). [9] R.J. Boland, et al., Knowledge representations and knowledge transfer, Academy of Management Journal 44 (2) (2001). [10] G.M. Campbell, Cross-utilization of workers whose capabilities differ, Management Science 45 (5) (1999 (May)). [11] P.R. Carlile, E.S. Rebentisch, Into the black box: the knowledge transformation cycle, Management Science 49 (9) (2003 (Sept)). [12] N.G. Carr, It doesnt matter, Harvard Business Review 81 (5) (2003 (May)). [13] J.E. Carrillo, G. Cheryl, Improving manufacturing performance through process change and knowledge creation, Management Science 46 (2) (2000). [15] T.H. Davenport, D.W.D. Long, M.C. Beers, Successful knowledge management projects, Sloan Management Review 39 (2) (1998 (Winter)).

T.S. Raghu, A. Vinze / Decision Support Systems 43 (2007) 10621079 [36] A. Majchrzak, L.P. Cooper, O.E. Neece, Knowledge reuse for innovation, Management Science 50 (2) (2004 (Feb)). [37] T.W. Malone, K. Crowston, The interdisciplinary study of coordination, ACM Computing Surveys 26 (1) (1994 (Mar)). [38] T.W. Malone, S.A. Smith, Modeling the performance of organizational structures, Operations Research 36 (3) (1988 (May/Jun)). [39] J.G. March, Bounded rationality, ambiguity, and the engineering of choice, Bell Journal of Economics 9 (2) (1978 (Autumn)). [40] H. Mintzberg, Structure in 5s: a synthesis of the research on organization design, Management Science 26 (3) (1980 (Mar)). [41] J.C. Moore, et al., Information acquisition policies for resource allocation among multiple agents, Information Systems Research 8 (2) (1997 (Jun)). [42] I. Nonaka, A dynamic theory of organizational knowledge creation, Organization Science 5 (1) (1994). [43] J.F. Nunamaker Jr., et al., Experiences at IBM with group support systems: a field study, Decision Support Systems 5 (2) (1989 (Jun)). [44] W.J. Orlikowski, Learning from notes: organizational issues in groupware implementation, Information Society 9 (3) (1993). [45] E.T. Penrose, The Theory of the Growth of the Firm, Wiley, New York, 1959. [46] M. Polanyi, Personal Knowledge: Toward a Post-Critical Philosophy, Harper Torchbooks, New York, 1962. [47] M. Polanyi, The Tacit Dimension, Routledge and Kegan Paul, London, 1966. [48] M.E. Porter, Industry structure and competitive strategy: keys to profitability, Financial Analysts Journal 36 (4) (1980 (Jul/Aug)). [49] C.K. Prahalad, R.M. Cowin, Developing strategic capability: an agenda for top management/comment, Human Resource Management 22 (3) (1983 (Fall)). [50] T.S. Raghu, A. Chaudhury, H.R. Rao, Business process change: a coordination mechanism approach, Knowledge and Process Management 5 (2) (1998). [51] H.R. Rao, A. Chaudhury, Modeling team processes: issues and a specific example, Information Systems Research 6 (3) (1995 (Sept)). [52] H.A. Simon, Administrative Behavior: A Study of Decisionmaking Processes in Administrative Organizations, 3rd Edition, Free Press, New York, 1976. [53] H.A. Simon, Prediction and prescription in systems modeling, Operations Research 38 (1) (1990 (Jan/Feb)). [54] B.L. Simonin, Ambiguity and the process of knowledge transfer in strategic alliances, Strategic Management Journal 20 (7) (1999). [55] The American Heritage Dictionary 2nd college ed. (Houghton Mifflin, Boston 1985). [56] J.D. Thompson, Organizations in Action, McGraw Hill, New York, 1967. [57] G.R. Ungson, D.N. Braunstein, P.D. Hall, Managerial information processing: a research review, Administrative Science Quarterly 26 (1) (1981 (Mar)).

1079

[58] Watts-Sussman, Adoption of Mediated Knowledge in Organizations: Source Credibility and Information Usefulness, (Unpublished Doctoral Dissertation, Boston University, 1998). [59] B. Wernerfelt, A resource-based view of the firm, Strategic Management Journal 5 (2) (1984). [60] M. Zeleny, Knowledge as a new form of capital: Part 1. Division and reintegration of knowledge, Human Systems Management 8 (1) (1989).

T.S. Raghu is Associate Professor of Information Systems in Arizona State University and a faculty associate of the Technology Research Center (CABIT). He received his PhD in Management Information Systems from SUNY Buffalo in 1999. His research interests are in Business Process Change, Knowledge Management and Collaborative Decision Making. He has also worked as a systems consultant at leading international IT consulting firms. His publications have appeared in refereed international journals such as Information Systems Research, Management Science, Journal of Organizational Computing and Electronic Commerce, Decision Support Systems and Expert Systems with Applications. A number of his papers have also appeared in the proceedings of refereed international conferences such as ICIS, AIS, and Informs.

Ajay Vinze is the Davis Distinguished Professor of Information Technology and the Director of the Technology Research Center (CABIT) at Arizona State University W.P. Carey School of Business. Prior to joining ASU, he served on the MIS faculty at Texas A&M University. He received his PhD in MIS from the University of Arizona, Tucson in 1988. Dr. Vinzes research, teaching and consulting interests focus on both IS strategy and technology issues. He has worked extensively on topics related to decision support technologies, knowledge management, computer supported collaborative work, and applications of artificial intelligence technology for business problem solving. His publications have appeared in leading MIS journals including Information Systems Research, MIS Quarterly, Decision Sciences, Journal of MIS and various IEEE Transactions . Dr. Vinze interfaces in various capacities with organizations in the USNASA, Intel, IBM, US Army ISEC, St. Lukes Episcopal Hospital, Arizona Department of Health Services, and internationally in Argentina, Australia, India, Mexico, New Zealand, Peru, Russia, Saudi Arabia and Trinidad. He is a member of INFORMS, Association of Information Systems and IEEE Computer Society.