KNOWLEDGE CREATION IN AN APPLICATION DOMAIN

KNOWLEDGE CREATION IN AN APPLICATION DOMAIN:
A HERMENEUTIC STUDY
TANYA LINDEN
Department of Management and Information Systems, Victoria University
P.O. Box 14428, Vic 8001, Australia
E-mail: tanya.linden@vu.edu.au
JACOB CYBULSKI
School of Information Systems, Deakin University
Burwood, Vic 3125, Australia
E-mail: jacob.cybulski@deakin.edu.au
In this paper we discuss a special case of knowledge creation (pattern mining) that was studied using a hermeneutic circle. Hermeneutics assists the oscillation between tacit and explicit knowledge through the process of knowledge qualification, combination, socialization, externalization, internalization and introspection. Our investigation of the knowledge creation process lead to the enrichment of the knowledge creation framework proposed by Wickramasinghe and Lichtenstein, allowing us to reflect on activities and aspects of knowledge elicitation across an application domain and involving practitioners who do not communicate directly.
1. Introduction
It is a common belief that organizational competitiveness depends on the company’s ability to acquire, accumulate and subsequently make effective use of their knowledge (Huber, 2001). Therefore one of the major dimensions of organizational knowledge management (KM) involves active employment of knowledge workers (ibid) since knowledge resides in people (Sutton, 2001). However, with the globalization of businesses KM researchers are focusing their attention on issues of knowledge management within and across organizations spanning across cultures, time zones, and geographical locations (Herbsleb & Mockus, 2003; Sarker & Sahay, 2004).
Past studies in software development (Mockus & Herbsleb, 2002; Orlikowski, 2002) on locating problem-solving expertise and sharing “best practices” explored the effects of distances and different cultural background on finding required expertise. However, in software development there is an approach - called pattern mining - that has been popular and proved itself useful in acquiring, recording, validating, enriching and sharing problem-solving knowledge and which may help overcome distances and cultural differences. Later this approach was transferred to other domains, e.g. organizational processes, education, human-computer interactions, etc.
In this study we examined the pattern mining process as a special case of knowledge management. We investigated how a facilitated approach can add value to finding domain experts and foster sharing of domain practices by recording them as patterns and pattern languages.

2. Background
The term pattern is used in various domains; however its meaning differs from one context to another, e.g. its meaning in art is different from the meaning of the same term in business operations. In the context of software development a pattern is defined as a literal form that allows recording a solution to a problem recurring in a certain context because it is being caused by certain forces (Schmidt, Johnson, & Fayad, 1996). When faced with a problem, developers often recall that they experienced something similar in the past and reuse the essence of the past approach to solve the new problem. Developers in different domains often view their experience as “problem-solution” pairs. When considering a solution it is necessary to take into consideration forces causing a problem in the first place. For example, when designing web sites web developers must take into consideration forces dealing with Internet technologies and forces related to the target audience of the website such as audience skill level. The solution must balance the conflict of the forces.
A pattern does not exist on its own. It applies to a particular domain and is connected to other patterns. Often to solve a problem a developer needs to apply a sequence of patterns. Such interrelated patterns form pattern languages (ibid.).
The pattern mining community, which aims to preserve problem solving knowledge in general, has a formal process that supports practitioners willing to share their knowledge or to elicit and represent knowledge of others in patterns. This pattern mining process is semi-formal. It consists of the following four stages (Manns, 2001):
(i) Pattern crafting – a person or a team willing to share their problem-solving experience records it in the format of patterns and pattern languages and submits this draft for shepherding (next stage of pattern mining).
(ii) Shepherding is an iterative process when the pattern authors make improvements to patterns in response to feedback received from the experienced pattern writer called a shepherd. Although it is desirable that a shepherd has knowledge of the domain for which patterns are written, it is not compulsory and therefore sometimes the main feedback given only reflects on pattern writing style but not on the problem-solving content.
(iii) A workshop session following shepherding is the stage where the patterns and pattern languages are evaluated by a group of experienced pattern writers. Like in the previous stage, although it is desirable to have workshop participants with the knowledge of the professional domain for which patterns are written, it is not a requirement.
(iv) Pattern improvement and dissemination is the final stage where authors update the patterns and pattern languages based on feedback received in the workshop and the patterns and pattern languages are made accessible to domain practitioners.

Pattern mining provides for capturing and dissemination of best practices and encourages continuous improvement of captured knowledge. Life of a pattern or a pattern language does not stop with its publication. Other practitioners who are interested in pattern writing can provide suggestions on pattern improvement. As time goes on, practitioners may get deeper understanding of the problem and its context, subsequently the pattern may be updated to reflect the changes (Rising, 1999). However, close examination of the process uncovered some deficiencies, especially in its first stage – pattern crafting (Linden & Cybulski, 2006). To participate in knowledge sharing, a practitioner has to learn pattern writing which is hard and time consuming. The pattern mining process in its widely adopted form relies blindly on the pattern authors’ judgements. Through all the stages of the process it is required that people evaluating the patterns (i.e. shepherds and workshop participants) are skilled in pattern writing, however their domain knowledge is desirable but not a requirement since “authors are experts” (Coplien & Woolf, 2000, p.561). As a result patterns and pattern languages may reflect the problem-solving experience of one person or a cohesive team of developers, rather than being a true representation of the accepted domain practices. Finally, for practitioners to be actively involved in pattern mining they have to be taken away from their working environment to attend pattern mining workshops and conferences which can be a problem for some practitioners.
Pattern mining can be viewed as a knowledge-creating process. Knowledge-creating processes are triggered by problem-solving (Leonard-Barton, 1995) and pattern mining aims to preserve problem solving knowledge. There exists a wide variety of models and frameworks of knowledge creation. Many of those are built upon one of the best-known and most-commonly used cognitive model SECI by Nonaka (1994). The SECI model describes transformation of knowledge forms during processes of socialization, externalization, combination and internalization. The model reflects the spiral nature of knowledge transition from individuals via various representations and its carriers to the level where it is shared as a collective knowledge whether by a group of individuals, or within an organization, or an inter-organizational domain.
The cyclical transitions described by Nonaka were explored by Wickramasinghe and Lichtenstein (2006) in a distributed knowledge creation context restricted by email exchange within an organizational setting. Wickramasinghe and Lichtenstein built their framework on the foundation of widely accepted models proposed by Nonaka (1994), Spender (1996) and Blackler (1995). The framework (see Fig. 1) describes the knowledge creation process with the first contributor sending an initial message by email to the micro-community. This “initial seed” is encoded since information is conveyed by signs and symbols and it is encultured since it is contributing to the process of developing shared understanding (Blackler, 1995). This first knowledge seed goes along a knowledge trail with its fragments being refined and / or additional fragments of knowledge being added along the way. This process is described as combination (adding knowledge fragments) and qualification (evaluating and refining). As a result integration of knowledge takes place where individual knowledge held by the participants of the process “has been combined, imbued with collective meaning and integrated into the group members’ tacit knowledge in a potentially valuable way” (Wickramasinghe & Lichtenstein, 2006). The entire process is self-managed by the participants and stops when the communication ceases. Every participant is directly involved in every cycle of knowledge combination and qualification.
Similar to knowledge creation in email, pattern mining is a knowledge-creating process and is cyclical in its nature, and therefore we examine it through the lens of the knowledge creation framework of Wickramasinghe and Lichtenstein (2006). In the initiation stage a practitioner-pattern writer (or a small team of writers) records the first draft of the pattern. Individual knowledge is externalized and transformed from tacit to explicit. This draft goes through the cycles of crystallization and sharing where combination and qualification occurs. First of all the authors of the draft contribute their knowledge and pattern writing skills to improve the draft. When they believe the draft is ready for the next stage, the shepherd is involved in combination-qualification cycles. The shepherd’s knowledge is also transformed from tacit to explicit and combined with authors’ knowledge. The resulting patterns are submitted to the next stage – workshop discussion – where the authors are given more suggestions on pattern refinement. Again in the workshop individual knowledge goes through tacit to explicit, individual to social transitions. As described earlier in the paper the pattern mining process has deficiencies concerning combination and qualification cycles. Combination is limited by the number of people involved and quite often it also impacts qualification since there is no procedure in place ensuring that patterns reflect acceptable domain practices – the process simply relies on the authors’ word for it.

Fig. 1. Knowledge creation framework (Wickramasinghe & Lichtenstein, 2006)
The Wickramasinghe and Lichtenstein framework (ibid.) helps understanding of knowledge creation between involved individuals; however, it has some limitations since it is based on the special case limited by boundaries of directly communicating individuals. It does not explore knowledge creation for a community of practitioners where contributors do not communicate directly and may not even know about each other and therefore may be coming from a much wider setting. At the same time, pattern mining although not providing any explanations for knowledge creation, supports capture, formalization and subsequent dissemination of practical knowledge among the developers who may benefit from the application of this knowledge and who are encouraged to contribute to creation of the knowledge.
Considering the benefits of pattern mining and challenges as well as benefits of domain-wide knowledge capture, formalization and sharing, this study investigated a special case of domain wide pattern mining. We explored the phenomena of facilitated knowledge creation involving domain practitioners who do not communicate directly and deal with a variety of application domain problem-solving tasks. The analysis of the case lead us to enriching the knowledge creation framework (Wickramasinghe & Lichtenstein, 2006) through the lens of our investigation.
3. Research design
Since this study is exploring knowledge creation across an application domain we needed to select such a domain that would be known for ample problem-solving cases incorporating both technological and non-technical issues. At the same time the domain should not be well-covered by pattern-writers but have good potential in presenting and sharing the problem-solving experience using patterns. The multimedia design field satisfied these criteria. It is known for continuous developments and for wide variety of problems that practitioners have been solving routinely in their ongoing work on projects. Also since modern applications use web browsers as their front-end and it is a trend for the future developments we decided that focusing on front-end multimedia design would give us an application domain with the well-defined boundaries, constant development potential and a wide range of practices coming from practitioners with diverse backgrounds.
Since we are not multimedia professionals, our research process needed to provide for deeper domain understanding including an examination of domain practices and exploration of multifaceted perspectives on problem-solving in that specific domain. Subsequently we expected the research data to be collected from domain practitioners and to be qualitative, rich, high in volume and convertible to text representation. For this study we were considering a qualitative research approach that would support collection of data in its context, analysis of rich data and most importantly, allowing flexibility in methods, enabling the analyst to follow new leads based on discoveries (Miles & Huberman, 1994). We needed to employ the research approach that was also interpretive, exploratory, and iterative.
The hermeneutic approach meets these criteria and has been known as focusing on human understanding, i.e. “understanding what people say and do, and why” (Myers, 2004, p.103). Hermeneutics, as applied in this study, was based predominantly on the work by Hans-Georg Gadamer (1976). There are five important notions central to Gadamerian hermeneutics which were exhibited by our study:
• Text. Hermeneutics provides a mode of analysis for textual data in search of its meaning (Myers, 1997). We were studying textual descriptions of pattern crafting (first stage of pattern mining) and texts derivatives as its outcome.
• Historicity and context. Hermeneutics interprets the text in terms of the context and time of its development. We studied problem-solving experience of practitioners in the context of undertaken projects as well as technological developments of the application domain, conducted during a specific period of time taking into consideration technology available at the time of the project implementation.
• Horizon of understanding and fusion of horizons. The concept of horizon of understanding was first introduced in 1930s by Husserl (1970) in his phenomenological theory where he states that each person has an individual understanding of phenomena. Understanding is a dynamic process and therefore “a horizon is not a rigid frontier, but something that moves with one and invites one to advance further” (Gadamer, 1979, p.217). The cycles of text interpretation bring together the horizons of the original text (or text analogue) and the horizon of the interpreter. “In the process of understanding there takes place a real fusion of horizons…” (Gadamer, 1979, p.273). The researcher’s role is to bring together the possible horizons on the studied issue as well as own horizons and pre-understanding. In this study we demonstrate fusion of horizons of domain practitioners, pattern writers and researchers.
• Hermeneutic circle. The process of understanding is iterative in its nature. By interpreting parts of a phenomenon and their relationships and then moving to better understanding of global context and then to an improved understanding of each part, the researcher goes through cyclical interpretations, referred to as the hermeneutic circle. This circular process continues until effective fusion of all horizons (Gadamer, 1979) consequently achieving a complete understanding of the investigated phenomena (Crotty, 1998; Olson & Carlisle, 2001). We studied the individual problem solving experiences of our subjects in order to understand the application domain. Through the iterative analysis and refinement of collected data we fused horizons of domain practitioners and pattern writers which resulted in multimedia design patterns.
• Interpreter’s prejudices and biases. Myers (2004) emphasizes that prejudice or prior knowledge is the initial point which drives the researcher’s understanding of the phenomena. Gadamer (1976) emphasized that it is natural for an individual to have prejudices since one may have past experiences with the phenomena under study and since it is impossible for a researcher to simply remove all preconceptions and presuppositions, such possible biases should be acknowledged and incorporated in the record of the observed phenomena. Having previous pattern mining experience and being intimately involved in this research, we had to identify our own preconceptions as well as our subjects’ biases and deal with them in each hermeneutic cycle.

Further developing Gadamer’s hermeneutics into a research method applicable for information systems (IS) Klein and Myers (1999) emphasized the importance of the cycles of mutual revision of meaning between parts and whole. In IS the hermeneutic circle has been applied to text analogues, such as an organization or an application domain. In such cases a field researcher may gain understanding of the selected domain as a whole through studying its parts, such as studying various documents, interviewing and observing practitioners in the domain, participating in their meetings, etc. (Myers, 2004).
In this study, interviews were selected as the main method for collecting data from practitioners. Hermeneutics is “doubly relevant to interview research, first by elucidating the dialogue producing the interview texts to be interpreted, and then by clarifying the subsequent process of interpreting the interview texts produced…” (Kvale, 1996, p.46). Interviewing can also be used in pattern crafting. For example, Rising (1999) used a combination of interviews and meetings to capture organizational experience which she subsequently represented in the pattern format.
Hermeneutics does not impose any restrictions on the number of sources involved in the hermeneutic circle. In IS research, as few as four (4) participants (Lee, 1994) have been used in hermeneutic investigations. In this study, in order to identify common experience across the selected domain, nine (9) participants representing the range of views and experiences were interviewed to achieve theoretical saturation (Strauss & Corbin, 1998). Hoffman et al. (1995) suggested initial unstructured interviews as a suitable approach to domain understanding with the follow-up interviews increasingly imposing the structure on the communication with practitioners.
Since hermeneutics employs iterative discovery of the subject matter, the exact steps and their number are determined based on the insights gained in the process. Therefore, it was impossible to state upfront the sequence and the number of cycles needed to achieve the goals of the study (Cole & O'Keefe, 2002). Instead in each hermeneutic cycle we continually re-assessed gained insights and our biases and presuppositions that could potentially influence our findings.
4. Empirical work
Through several hermeneutic cycles (see details in Linden, 2007; Linden & Cybulski, 2006) we studied knowledge creation as represented by the pattern mining process in the area of front-end multimedia design. The major attention has been paid to the pattern crafting stage of the process since this stage caused major concerns and exhibited deficiencies described in section 2. During the study we refined and partially evaluated the pattern crafting stage. In doing so we generalized our findings and reflected them in the knowledge creation framework of Wickramasinghe and Lichtenstein (2006). The resulting knowledge artifacts – the multimedia design patterns have been reported elsewhere (Linden & Cybulski, 2005).
4.1. First hermeneutic cycle – studying the existing pattern mining process
The studied text included publications on patterns and pattern mining. As a result of the study the deficiencies of the process mentioned earlier in this paper have been identified.
4.2. Second hermeneutic cycle – understanding the application domain
The domain understanding started with externalizing experience of domain practitioners. The text to study contained the transcribed interviews. Design issues in the multimedia development process that were important to practitioners were identified and a fine-grain textual system of codes has been developed. This rich coding reflected both the historicity and the context of projects and represented expansion of the researchers’ horizon and its gradual fusion with practitioners’ horizons. Domain understanding for the researchers resulted in internalization of the acquired knowledge. The developed codes laid the foundations for a conceptual model of domain knowledge as communicated by the interviewees.
We recognized that rich understanding of the multimedia domain gained in this round could influence our subsequent work. We recognized it as a threat of possible bias which had to be controlled, especially at the time of producing patterns.
4.3. Third hermeneutic cycle – producing pattern skeletons
In this cycle the researchers interpreted original interview texts as well as the system of codes developed in the previous cycle. The analysis of the texts allowed design of the patterns skeletons – a step towards formalizing collected problem-solving experience. The resulted patterns (although with gaps) represented our horizon of understanding of domain practices. These formalized representations were fed as the first seeds for the combination and qualification cycles as described by the knowledge creation framework.
4.4. Fourth hermeneutic cycle – filling in the gaps
The fourth cycle of this study aimed at investigating the feasibility of producing full patterns. Two approaches were trialed.
• In the first approach we conducted a focus group session, like a mini pattern-mining workshop, where experienced pattern writers and multimedia developers discussed the pattern skeletons created in the previous cycle. The process of participants’ socialization and externalization of their knowledge resulted in one full pattern (due to time limitations) which represented full fusion of horizons of domain practitioners and pattern writers.
• In the second approach individual multimedia designers were approached with the request to provide missing details in order to fill in the gaps of the pattern skeletons. In three separate sessions the externalized knowledge of practitioners was combined and formalized into full patterns also representing fusion of horizons of all contributing domain practitioners and the pattern writer.

In this cycle the researchers acting as facilitators had to control their pre-understanding of the subject matter and their bias in order to represent only externalized knowledge of the multimedia developers and pattern writers.
5. Discussion
The hermeneutic circle consisting of the cycles described in the previous section aimed at introspective reflection on the emerging process of pattern crafting. As part of this introspection, after each hermeneutic cycle was completed, the researchers recorded the changes to the pattern crafting process (as reported in Linden & Cybulski, 2006), and discussed how their insights should be reflected in the original knowledge creation framework (Wickramasinghe & Lichtenstein, 2006).
The initial seed fed in the process reflects the externalization of problem solving knowledge contributed by domain practitioners. This knowledge has not been through any combination or qualification yet, it is raw data representing individual experience obtained through transformation from tacit to explicit form (externalization as defined by Nonaka (1994)). This knowledge coming from individuals is being examined for problems common to the domain. Subsequently suggested solutions are qualified in terms of their acceptance by domain practitioners (combination and qualification in the process of socialization).

Fig. 2. Knowledge creation framework reflecting our study

To reflect the context of this project and its hermeneutic approach in the knowledge creation framework, four sections have been defined (see Figure 2): domain knowledge/experience, formalized experience (patterns), pattern mining process (the process for knowledge combination and qualification) and introspection (to note presuppositions and biases of involved participants). These sections correspond to the points emphasized by the knowledge creation framework of Wickramasinghe and Lichtenstein (ibid.), i.e. tacit knowledge is represented by the domain knowledge and experience section; explicit knowledge is the patterns and pattern languages section depicting formalized experience; pattern mining process involves combination and qualification of knowledge and it also reflects the social dimension of knowledge whereas introspection reflects the individual dimension. The spiral shows that the four dimensions are being revisited in the true spirit of hermeneutics. On this diagram we also showed where our four hermeneutic cycles appear in the spiral (depicted by numbers 1-4). During the focus group session with pattern writers and domain practitioners evaluation of patterns happened alongside formalization, and therefore evaluation is also depicted on the spiral.

Fig. 3. Knowledge creation framework
As a result of this study, we examined pattern mining as a special case of knowledge creation (see Figure 3). Knowledge creation is a process where individual knowledge of contributors is externalized reflecting horizons of understanding of participants of the pattern mining process. Through the cycles of formalization and evaluation the full fusion of horizons takes place.
Since the result pattern mining process is patterns publication leading to sharing and application of knowledge, we added these phases to the spiral. It has been noted that the original knowledge creation framework (Wickramasinghe & Lichtenstein, 2006) stemmed from the work on knowledge development lifecycle which included application as its phase (Lichtenstein, 2004).
We had to consider two aspects of this study: one – the pattern mining process or in general terms knowledge creating activities, another – the outcome of the studied process, i.e. knowledge itself (in our case presented in the format of patterns and pattern languages). In the resulting framework the knowledge creation process leads from individual understanding of problem-solving to the facilitated formalization and evaluation of this knowledge, and the subsequent sharing and application (by individual practitioners or in project teams). Individual and group contribution of knowledge results in oscillations between tacit and explicit knowledge. In the depicted cycles knowledge combination and qualification contributes to knowledge crystallization and sharing. In addition the resulting framework shows the dimension of individual introspection which reflects on participants understanding of their role in the knowledge creation process and their prejudices that may influence their problem solving activities.
6. Summary and conclusions
In this paper we explored the possibility of integrating and enriching two systems of concepts: the knowledge creation framework proposed by Wickramasinghe and Lichtenstein (Wickramasinghe & Lichtenstein, 2006) and the pattern mining process as currently applied by pattern mining communities all over the world (Manns, 2001). At the end of this study:
• we gained rich understating of the selected application domain;
• we succeeded in recording selected problem-solving experience in the format of patterns and pattern languages;
• we proposed a formalized facilitated approach to the pattern crafting stage of pattern mining as emerged from our study;
• we recorded reflections of researchers on the way they influenced the research investigation; and
• we enriched the knowledge creation framework with our insights.

The resulting knowledge creation framework reflects the spiral essence of knowledge creation and its oscillations between various forms (tacit-explicit, internal-external, individual-social). In addition to the original framework there is a clear representation of two aspects: knowledge and process (activities resulting in knowledge). By applying hermeneutics we added a dimension of introspection which was missing from the original framework.
As any other interpretive research, the reported study is limited in its ability to generalize findings (Sarkar & Cybulski, 2004). In our future research we would like to investigate the transferability of the discovered process from the multimedia domain to other application domains.
References
Blackler, F. (1995). Knowledge, Knowledge Work and Organizations: An Overview and Interpretation. Organization Studies, 16(6), 1021-1046.
Cole, M., & O'Keefe, R. M. (2002). Hermeneutic Philosophy and Data Collection: A Practical Framework. Paper presented at the Eighth Americas Conference on Information Systems.
Coplien, J. O., & Woolf, B. (2000). A Pattern Language for Writers' Workshops. In N. Harrison, B. Foote & H. Rohnert (Eds.), Pattern Languages of Program Design (Vol. 4, Chapter 25, pp. 557-583). Reading, MA: Addison-Wesley.
Crotty, M. (1998). The Foundations of Social Research. Crows Nest, NSW: Allen & Unwin.
Gadamer, H.-G. (1976). Philosophical Hermeneutics. Berkeley, California: University of California Press.
Gadamer, H.-G. (1979). Truth and Method. London: Sheed and Ward.
Herbsleb, J. D., & Mockus, A. (2003). An Empirical Study of Speed and Communication in Globally-Distributed Software Development. IEEE Transactions on Software Engineering, 29(6), 1-14.
Hoffman, R. R., Shadbolt, N. R., & Burton, A. M. (1995). Eliciting knowledge from Experts: A Methodological Analysis. Organisational Behavior and Human Decision Processes, 62(2), 129-158.
Huber, G. P. (2001). Transfer of Knowledge in Knowledge Management Systems: Unexplored Issues and Suggested Studies. European Journal of Information Systems, 10(2), 72-79.
Husserl, E. (1970). Logical Investigation. Atlantic Highlands, NJ: Humanities Press.
Klein, H. K., & Myers, M. D. (1999). A Set of Principles for Conducting and Evaluating Interpretive Field Studies in Information Systems. MIS Quarterly, 23(1), 67-93.
Kvale, S. (1996). InterViews. An Introduction to Qualitative Research Interviewing. Thousands Oaks, California, USA: Sage Publications.
Lee, A. S. (1994). Electronic Mail as a Medium for Rich Communication: An Empirical Investigation Using Hermeneutic Interpretation. MIS Quarterly, 18(2), 143-157.
Leonard-Barton, D. (1995). Wellsprings of Knowledge: Building and Sustaining the Sources of Innovation. Boston, Massachusets: Harvard Business School Press.
Lichtenstein, S. (2004). Knowledge Development and Creation in Email. Paper presented at the 37th Hawaii International Conference on System Sciences (HICSS 2004), Hawaii, USA.
Linden, T. (2007). Acquiring Problem-Solving Experience in the Multimedia Domain. Paper presented at the IRMA 2007.
Linden, T., & Cybulski, J. L. (2005). Colour Issues in Web Development. Paper presented at the 10th European Conference on Pattern Languages of Programs (EuroPLOP 2005), Irsee, Germany.
Linden, T., & Cybulski, J. L. (2006). Refining the Process of Sharing Problem-Solving Experience across Domain: A Hermeneutic Study. Paper presented at the American Conference on Information Systems AMCIS 2006, Acapulco, Mexico.
Manns, M. L. (2001). Patterns: A Promising Approach to Knowledge Management. Paper presented at the First Annual ABIT Conference, Pittsburgh, Pennsylvania, USA.
Miles, M. B., & Huberman, A. M. (1994). Qualitative Data Analysis: An Expanded Sourcebook (Second ed.). Thousands Oaks, California: Sage Publications.
Mockus, A., & Herbsleb, J. D. (2002). Expertise Browser: A Quantitative Approach to Identifying Expertise. Paper presented at the International Conference on Software Engineering, Orlando, Florida.
Myers, M. D. (1997). Qualitative Research in Information Systems. MIS Quarterly, 21(2), 241-242.
Myers, M. D. (2004). Hermeneutics in Information Systems Research. In J. Mingers & L. P. Willcocks (Eds.), Social Theory and Philosophy for Information Systems (pp. 103-128). Chichester: John Wiley & Sons.
Nonaka, I. (1994). A Dynamic Theory of Organizational Knowledge Creation. Organizational Science, 5(1), 14-37.
Olson, D. L., & Carlisle, J. (2001). Hermeneutics in Information Systems. Paper presented at the Seventh Americas Conference on Information Systems, Boston, Massachusetts USA.
Orlikowski, W. J. (2002). Knowing in Practice: Enacting a Collective Capability in Distributed Organizing. Organization Science, 13(3), 249-273.
Rising, L. (1999). Patterns Mining. In S. Zamir (Ed.), Handbook of Object Technology (pp. 38-31 - 38-39 (Chapter 38)). New York: CRC Press.
Sarkar, P., & Cybulski, J. L. (2004). Evaluation of Phenomenological Findings in IS Research: A Study in Developing Web-Based IS. Paper presented at the Twelfth European Conference on Information Systems (ECIS 2004), Turku School of Economics and Business Administration, Turku, Finland.
Sarker, S., & Sahay, S. (2004). Implications of Space and Time for Distributed Work: An Interpretive Study of US-Norwegian System Development Teams. European Journal of Information Systems, 13(1), 3-20.
Schmidt, D. C., Johnson, R. E., & Fayad, M. (1996). Software Patterns. Communications of the ACM, 39(10 - Special Issue on Patterns and Pattern Languages), 37-39.
Spender, J. C. (1996). Making Knowledge the Basis of a Dynamic Strategy of the Firm. Journal of Strategic Management, 17(Winter), 45-62.
Strauss, A., & Corbin, J. (1998). Basics of Qualitative Research. Grounded Theory Procedures and Techniques (Second edition ed.). Thousand Oaks, CA, USA: Sage Publications.
Sutton, D. C. (2001). What Is Knowledge and Can It Be Managed? European Journal of Information Systems, 10(2), 80-88.
Wickramasinghe, N., & Lichtenstein, S. (2006). Supporting Knowledge Creation with E-mail. International Journal of Innovation and Learning, 3(4), 416-426.