A New Methodology for Distributed Knowledge Management Analysis Roberta Cuel (Department of Computer and Management Sciences University of Trento, Italy
[email protected])
Abstract: The Distributed Knowledge Management (DKM) approach tries to overcome problems deriving from a typical outcome of traditional Knowledge Management (KM) solutions: the creation of a unique conceptualization of corporate knowledge (e.g. an ontology or a unique system of classification) which does not allow autonomy of organizational units and the coexistence of different conceptual schemas (e.g. points of view or perspectives). This new approach takes strongly into account local heterogeneity, and looks at complex knowledge-based organizations as constellations of local organizational units which manage knowledge in an autonomous way, exchanging it with other units through meaning negotiation/coordination processes. In a DKM system each autonomous unit is reified by a Knowledge Node (KN), a useful abstraction which allows us to identify, within organizations, people who manage knowledge according to a local conceptual schema and a personalized system of artifacts. In this paper, a new methodology is described and then used to unveil Knowledge Nodes in a case study: Impresa Pizzarotti & C. S.p.A., a complex Italian building industry. It is argued that the resulting Distributed Knowledge Management system might be implemented within the firm with a high probability of success, because the system of Knowledge Nodes (discovered through this methodology) reflects the perspective and the way in which organizational units usually manage their knowledge. Key Words: Distributed Knowledge Management, Knowledge Nodes, Autonomy, Coordination, Methodology for Organizational Analysis. Category: C.2.4, H.1.0
1
Introduction
Most Knowledge Management (KM) systems aim at creating large, homogeneous knowledge repositories, in which corporate knowledge is made explicit, unified, represented and organized according to a unique – supposedly shared – conceptual schema. The common outcome of KM systems is the creation of an Enterprise Knowledge Portal (EKP), a (web-based) interface which provides a common access point to corporate knowledge. Even if users have different profiling systems, the underlying representation of EKP is typically unique, and is meant to represent a common and shared conceptualization of corporate knowledge that enables communication and knowledge sharing across the entire organization [Davenport et al., 1998], [Borghoff and Pareschi, 1998]. In this work a new approach – called Distributed Knowledge Management (DKM) approach [Bonifacio et al., 2000], [Bonifacio et al., 2002b] – is used. It is
based on the belief that traditional KM systems, and in particular the unique and objective conceptualization of corporate knowledge, are incompatible with the very nature of what is to be managed (i.e., knowledge), and consequently are often deserted [Bonifacio et al., 2002d] by users. This approach thoroughly analyzes subjective and social aspects of knowledge which are considered irreducible to a common, unique and completely sharable conceptualization and is thought of as local [Ghidini and Giunchiglia, 2001], [Benerecetti et al., 2000], [Goffman, 1974], [Fauconnier, 1985], [Dinsmore, 1991], and the result of social negotiations (individual participation and objects reification [Wenger, 1998]) within organizational units. The DKM approach is based on two main principles: the principle of autonomy, which grants organizational units a high degree of semantic autonomy in managing their local knowledge; and the principle of coordination, which allows each organizational unit to exchange knowledge with other units through processes of double loop learning [Argyris, 1999] or perspective taking [Boland and R.V.Tenkasi, 1995]. According to this approach, complex knowledge-based organizations can be seen as “constellations” of local organizational units which exhibit some degree of semantic autonomy, namely the ability to manage locally knowledge and to develop a personal perspective on the world. The resulting KM systems should sustain the creation and management of different conceptual schemas which coexist within a DKM system.
2
Knowledge Node: the building block of DKM
Within a DKM system, each organizational unit – either formal (e.g. divisions, market sectors) or informal (e.g. interest groups, communities of practice, communities of knowing) – must be represented and reified by a Knowledge Node (KN) [Bonifacio et al., 2002a]. As shown in figure 1(a), each KN represents a semantically autonomous unit composed by: – a knowledge owner: an individual or a collective entity that has the capability of managing its own knowledge (e.g. worker, manager, team, community, office); – a system of artifacts: the system of procedures, activities, documents, archives, technologies, languages, and so on, used by the knowledge owner to manage local knowledge in a way that best suits its environmental condition and its needs; – a context: an explicit representation of a knowledge owner’s perspective (or personal conceptual schema).
(a) Knowledge Node
(b) DKM system
Figure 1: A DKM system composed by KNs
Therefore, a DKM system (in figure 1(b)) will be composed by a “constellation of KNs”, each managing knowledge in an autonomous way, and exchanging it with others, without sharing a common and unique conceptualization of knowledge, but through negotiations/coordination processes, and boundary encounters managed by brokers (human or not) [Bonifacio et al., 2002c]1 .
3
A New Methodology of Analysis
In order to develop a DKM system within the firm, an effective methodology of analysis is needed to focus the attention on two relevant aspects based on the two DKM’s principles: – identify the borders of existing KNs within the firm, namely organizational units which exhibit some degree of semantic autonomy (principle of autonomy); – identify the way in which knowledge is exchanged across the whole organization through negotiation/coordination processes (principle of coordination). Both are based on social relations within and across communities in the firm; and these connections can be analyzed using different methodologies such as social network analysis (SNA) [Wasserman et al., 1994], I* [Yu, 1997], Tropos [Castro et al., 2002], ethnography [Orlikowski and Gash, 1994], and so on. SNA 1
This architecture is under development as part of EDAMOK (http://edamok.itc.it/), a joint project carried out by the Institute for Scientific and Technological Research (IRST, Trento) and by the University of Trento (Italy).
and other quantitative methodologies give us a good and general perspective on the organization and allow us to perceive the real structure of the organizational model by considering the relations among people and groups, but do not allow us to identify the reason why some groups are strategic and others are not. On the other hand, ethnography and other qualitative methodologies are based on the participation of the observer within the firm, who tries to get a detailed understanding of the circumstances of few subjects being studied but cannot determine the significance of what she/he observes without gathering broad statistical information. In this work it has been decided that these kinds of analysis are not enough to figure out KNs [Cuel, 2003], since it is difficult to identify their boundaries and their knowledge exchanging processes. As a matter of fact, individuals belonging to an organizational unit are socially interconnected to solve different objectives and are often part of two or more units, thus using more than one conceptual schema. It was decided to develop a series of questionnaires and a sort of ethnographic interviews [Spradley, 1979], in order to: 1. understand the main picture of the firm: to understand what happens within the firm, which kind of procedure are developed, the nature of coexistent relations and communications among organizational units, and so on; 2. unveil KNs and their relations: to reify formal or informal organizational units which exhibit some degree of semantic autonomy, and the way KNs communicate and exchange knowledge. In particular it is necessary to analyze: – the knowledge owner, which is one or more individuals who frequently interacts in an informal way adapting and managing their local knowledge by developing a common conceptual schema; – the system of artifacts, used by knowledge owners, that expresses the local knowledge and a KN’s semantic autonomy; some other elements (such as specific language, way of thinking) which reflect a common identity; – a shared conceptual schema resulting from the continuous processes of meaning negotiation/coordination among knowledge owners; – the way in which knowledge owners (human or not) allow processes of meaning negotiation/coordination among KNs through boundary objects, brokers, and boundary encounters [Wenger et al., 1994]; 3. validate the first results through focus groups, or meetings with workers involved in the organization activity.
Even if interviewers have to figure out these elements – making efforts on specific aspects –, the style of ethnographic interview is completely free, allowing trust and comfortable discussions between the interviewer and the person interviewed. This methodology has been applied to a complex Italian building industry, the Impresa Pizzarotti & C. S.p.A. The aims of the analysis were: – to develop a high level architecture of KM systems based on the DKM approach, unveiling KNs and their relations; – to figure out which kind of knowledge can be managed within an autonomous organizational unit and how it can be made available to others. Pizzarotti & C. S.p.A. has a complex organizational model because of the nature of the product and the characteristics of the sector. Production processes need to be localized in the areas in which the product is created, therefore building yards are spread out over the country and in Europe. The central offices are organized as services which support the production (e.g. the security office, the quality office, the administration office), and each service has its own organization and structure. To unveil KNs we made ethnographic interviews with 36 workers selected in different offices and roles in the firm, and we visited 3 different building yards. This analysis allowed us to unveil KNs looking at knowledge owners, the systems of artifacts, the contexts and, more importantly, the kind of knowledge that is exchanged within a community and the way in which people negotiate/coordinate knowledge across the whole organization. Each building yard must solve specific problems which are connected with the kind of production, and other environmental factors (e.g. the weather, local costumers and suppliers); and each service office has its own structure and way of working which expresses semantic autonomy. Therefore, local building yards and centralized offices have been reified as KNs. Moreover, in this analysis some other units, which exhibit some degree of semantic autonomy, have been discovered: the group of project managers and the group of yard managers. These groups are not formally recognized by the firm, and are composed by workers (belonging to several other units) who need to manage knowledge according to a particular conceptual schema, and to use a personalized system of artifacts in order to work. For instance, the project managers communicate all together in an informal way (e.g. phone-calls, discussions over a cup of coffee) to develop common knowledge and a shared conceptualization schema that are used to develop new projects; they share systems of artifacts (such as specific timetables) and a specific language (such as technical words) which allow them to manage their work. Finally, we proposed our results to workers and other people responsible of the project to validate the idea of the two different KNs not formally recognized
within the firm. It has been figured out that different strategic but informal KNs coexist within the firm in order to solve common problems, that other formal units are not able to manage.
4
Conclusion
In this work, we have focused the attention on the lack of coherence between a privileged, unique, and supposedly shared conceptualization of knowledge within a KM system, and the different ways of thinking of workers, communities, teams, offices that participate in the firm activity. This could be considered as one of the reasons why people are led to desert KM systems in that they consider them either oppressive or irrelevant [Bowker and Star, 1999]. From a theoretical point of view, and through the analysis of Impresa Pizzarotti & C. S.p.A. – and other case studies such as [Bonifacio et al., 2000], [Benassi and Greve, 1996] –, it has been proved that the resulting DKM system might be implemented within the firm with a high probability of success. In other words, it will not be perceived by workers as oppressive or irrelevant [Bowker and Star, 1999], because KNs reflect the way in which organizational units usually manage their knowledge. Finally, considering the organization as a dynamic organism, the DKM system has to be very flexible and has to set the stage for the introduction of new KNs and the modification of others. The future work will address a study of effectiveness of DKM systems within firms. In particular, there are some organizations which might adopt the DKM system – developed in the project EDAMOK – and it will be very interesting to measure the real effectiveness of both the methodology of analysis of KNs and the implementation of DKM systems.
References [Argyris, 1999] Argyris, C. (1999). Teaching Smart People How to Learn. Harvard business review. [Benassi and Greve, 1996] Benassi, M. and Greve, A. (1996). GESTO– A Network Organized Company? Journal of Market Focused Management, 1:301–323. [Benerecetti et al., 2000] Benerecetti, M., Bouquet, P., and Ghidini, C. (2000). Contextual Reasoning Distilled. Journal of Theoretical and Experimental Artificial Intelligence, 12(3):279–305. [Boland and R.V.Tenkasi, 1995] Boland, J., and Tenkasi, R.V. (1995). Perspective Making and Perspective Taking in Communities of Knowing. Organization Science, 6(4):350–372. [Bonifacio et al., 2002a] Bonifacio, M., Bouquet, P., and Cuel, R. (2002a). Knowledge Nodes: the Building Blocks of a Distributed Approach to Knowledge Management. Journal of Universal Computer Science, 8(6):652–661. Springer Pub. & Co. [Bonifacio et al., 2002b] Bonifacio, M., Bouquet, P., and Cuel, R. (2002b). The Role of Classification(s) in Distributed Knowledge Management. In Proceedings of 6th International Conference on Knowledge-Based Intelligent Information Engineering Systems & Allied Technologies (KES’2002), Amsterdam, the Netherlands. IOS Press.
[Bonifacio et al., 2002c] Bonifacio, M., Bouquet, P., Mameli, G., and Nori, M. (2002c). Kex: a Peer-to-Peer Solution for Distributed Knowledge Management. In Karagiannis, D. and Reimer, U., editors, Fourth International Conference on Practical Aspects of Knowledge Management (PAKM-2002), Vienna (Austria). [Bonifacio et al., 2000] Bonifacio, M., Bouquet, P., and Manzardo, A. (Stanford, March 18-21, 2000). A Distributed Intelligence Paradigm for Knowledge Management. In AAAI Spring Symposium Series 2000 on Bringing Knowledge to Business Processes. AAAI. [Bonifacio et al., 2002d] Bonifacio, M., Bouquet, P., and Traverso, P. (2002d). Enabling Distributed Knowledge Management. Managerial and Technological Implications. Novatica and Informatik/Informatique, III(1). [Bonifacio et al., 2002e] Bonifacio, M., Cuel, R., Mameli, G., and Nori, M. (8-10 October 2002e). A Peer-to-Peer Architecture for Distributed Knowledge Management. In Proceedings of 3rd International Symposium on Multi-Agent Systems, Large Complex Systems, and E-Businesses (MALCEB’2002)., Erfurt/Thuringia, Germany. [Borghoff and Pareschi, 1998] Borghoff, U. M. and Pareschi, R. (1998). Information Technology for Knowledge Management. Springer. [Bowker and Star, 1999] Bowker, G. C. and Star, S. L. (1999). Sorting Things Out: Classification and Its Consequences. MIT Press. [Castro et al., 2002] Castro, J., Kolp, M., and Mylopoulos, J. (2002). Towards Requirements-Driven Information Systems Engineering: the TROPOS Project. Information Systems. [Cuel, 2003] Cuel, R. (2003). A Knowledge-Oriented Analysis of Organizations. Knowledge Nodes: the Building Blocks of a Distributed Knowledge Management Approach. Proceedings of Doctoral Consortium (DCEIS 2003)., Angers – France. [Davenport et al., 1998] Davenport, T. H., Long, D. W. D., and Beers, M. C. (1998). Successful Knowledge Management Projects. Sloan Management Review, 39(2). [Dinsmore, 1991] Dinsmore, J. (1991). Partitioned Representations. Kluwer Academic Publishers. [Fauconnier, 1985] Fauconnier, G. (1985). Mental Spaces: aspects of meaning construction in natural language. MIT Press. [Ghidini and Giunchiglia, 2001] Ghidini, C. and Giunchiglia, F. (2001). Local Models Semantics, or Contextual Reasoning = Locality + Compatibility. Artificial Intelligence, 127(2):221–259. [Goffman, 1974] Goffman, I. (1974). Frame Analysis. Harper & Row, New York. [Orlikowski and Gash, 1994] Orlikowski, W. J., and Gash, D. C. (1994). Technological Frames: Making Sense of Information Technology in Organization. ACM Transactions on Information Systems, 12(2):174–207. [Spradley, 1979] Spradley, J. P. (1979). The Ethnographic Interview. Holt, Rinehart, Winston, New York. [Wasserman et al., 1994] Wasserman, S., Faust, K., and Iacobucci, D. (1994). Social Network Analysis : Methods and Applications. Cambridge University Press. [Wenger, 1998] Wenger, E. (1998). Communities of Practice. Learning, Meaning, and Identity. Cambridge University Press. [Wenger et al., 1994] Wenger E. C. and Snyder W. M. (2000). Communities of Practice: The Organizational Frontier. Harvard Business Review, 139–145. [Yu, 1997] Yu E. (Washington D.C., 226-235, 1997). Towards Modelling and Reasoning Support for Early-Phase Requirements Engineering. In Proceedings of the 3rd IEEE Int. Symp. on Requirements Engineering.
