2nd EUROPEAN COMPUTING CONFERENCE (ECC’08) Malta, September 11-13, 2008

Capturing Software- Tacit

ABDULMAJID HISSEN MOHAMED Department of Computer Sebha University PO Box 18758, Sebha LIBYA [email protected] http://www.sebhau.edu.ly/cs/faculty.html

Abstract: - systems are aimed to provide knowledge-intensive organisations with tools, and methods to better manage their knowledge capital. A great success is gained in the course of managing explicit knowledge in the form documented knowledge fragments. But, greater part of organisations knowledge is realised in tacit form which is volatile and hardly captured in a formal way. Managing this type of knowledge still represents one of the major challenges in knowledge management research. This paper proposes a knowledge model that caters for capturing both tacit and explicit organisational knowledge in the software-engineering domain.

Key-Words: - Knowledge Management, software, tacit knowledge, knowledge reuse, software artefacts

1 Introduction management. Few attempts have been made to As a result of the emerging knowledge-based manage or at least partially capitalise tacit economy (or K-Economy), knowledge has become knowledge, but these efforts suffer limitations in the strategic resource of organizations to compete managing it adequately. Capturing tacit knowledge and survive. In line with this trend, the introduction is more complex than the management of explicit of knowledge management systems (KMS) is knowledge for multiple reasons. For example, tacit receiving greater attention. KMS aims at providing knowledge is highly personal and organisations have an organization the ability to benefit from its past no control on its sharing and exploitation. Also there in order to respond more effectively in are many psychological as well as technical barriers present situations. Organizational knowledge could that work against capturing the full richness of tacit be distinguished as formal (explicit) and informal knowledge. (tacit) knowledge. According to Conklin [4], Formal The rest of the paper is organised as knowledge is the stuff of books, manuals, documents, follows: first, the importance of capturing tacit memos, white papers, plans and training courses, knowledge is highlighted. Second, an overview of whereas informal knowledge is the knowledge that is technical and non technical challenges of capturing created and used in the process of creating the tacit knowledge is discussed. Third, the conceptual formal results. It includes ideas, , assumptions, framework and components of the proposed model questions, guesses, stories, points of view, etc. is explained. The paper ends with a conclusion and Notice that these forms of tacit knowledge are suggestion for further work. personal and only reside in experts’ heads. In short, it constitutes what Koskinen [8] describes as the practical know-how. 2. The importance of tacit knowledge Traditional KMS has mostly focused on the Experience is an indispensable resource for any collection and dissemination of explicit (formal) organization to perform better in handling upcoming knowledge, which is structured and stored in similar situations. Unfortunately organizations do electronic knowledge repositories. But, there is a not control all knowledge assets they have. Great growing recognition that tacit knowledge constitutes portion of organizational knowledge is realized in an essential part of the organisational memory, and tacit form which usually resides in workers’ heads it has to be considered as well. According to Zack as personal belief (know-how). Unfortunately when [15], explicating tacit knowledge so it can be a human expert becomes unavailable either efficiently and meaningfully shared and reapplied, temporarily or permanently, then his/her expertise especially outside the originating community, is one becomes unavailable too. In many situations of the least understood aspects of knowledge knowledge workers who have been ousted for

ISSN:1790-5109 132 ISBN: 978-960-474-002-4 2nd EUROPEAN COMPUTING CONFERENCE (ECC’08) Malta, September 11-13, 2008

various reasons, are called back for consultation. status, laziness or controversial ideas” [1]. All these This situation simply occurs because part of the problems have its roots in human psychological organisation’s knowledge is taken away and it has to behaviour. Therefore, we believe that the be retained. In a worst scenario, the knowledge externalisation of tacit knowledge has to be carried expert might move to one of the business out as part of a defined knowledge lifecycle. It competitors, which means partially disarming should not be left as merely voluntarily act. For knowledge of the former employer. This is why in example, in software engineering, deliberations held organizations where workers turnover is high such in different reviews of software projects lifecycle as software engineering, the codification of tacit including the post-mortem reviews have to be knowledge becomes an extremely important issue. formally captured. Unfortunately, the importance of The importance of managing tacit knowledge is well the integration between software process models and expressed in the following quote by Cummings [3], knowledge lifecycle has probably not yet been “An organization’s knowledge walks out of the door sufficiently understood. every night – and it might never come back”. Unfortunately, recent huge investment in IT is focused more on the management of explicit 3.2 Technical Obstacles knowledge with less emphasis on the part of tacit In addition to non technical obstacles that contribute knowledge management. Johannessen et al [6] to difficulties in recognising and capturing tacit asserts that companies easily lose their competitive knowledge, tacit knowledge is usually less edge if they emphasis investment in, and use of IT structured and it does not conform to a standard without taking tacit knowledge into consideration. format. Therefore, it can be hardly coded and captured using traditional data models. From the software engineering viewpoint, 3 Barriers to capturing tacit models such as DRL [9], QOC [11] and IBIS [13] knowledge are all aimed to codify tacit knowledge in the form of different alternatives and justifications behind 3.1 Non-Technical obstacles developers’ design decisions. In spite of the support Knowledge management is not a technology of capturing tacit knowledge as facilitated by these problem [14, 10]. Since it involves the models, they cannot accommodate the full richness externalisation of human expertise, human factors of tacit knowledge. Sometimes, people are expected highly influence the success of any KMS. These to articulate their tacit knowledge differently; they non-technical factors are mostly related to firstly, might even resort to the use of body . There how to characterise human knowledge, and is no one model that can accommodate all these secondly, how to promote knowledge workers to variations in expressing tacit knowledge. Another proactively contribute to externalise their tacit major limitation shared by all of these models is knowledge in order to build a group memory that, they are not coupled well with explicit system. Usually there is a tendency that knowledge knowledge in the form of knowledge embedded workers are very reluctant to share their personal artefacts. We believe knowledge artefacts should be knowledge with others. Some workers regard their regarded as incubators of tacit knowledge. professional knowledge as a source of job security The widespread use of the Internet made and they are often reluctant to share it. Other email systems an easy way to capture tacit organisational practices might also contribute to knowledge. This approach can be exemplified by workers reluctance to share knowledge. According TeamInfo[2]. But, this approach still has limitations to Kokkoniemi [7], in organisations where staff can in the form of the lack of face to face interactions be rewarded for having particular expertise, this where richer tacit knowledge is likely to surface. might lead to staff do not want to share what they Another major problem of email-based approaches know. In addition, psychological problems might lies in the that, the captured knowledge (in the also influence the position of knowledge workers form of email messages) is not sufficiently participation in KM settings. For example, in accessible by computational means because they do meetings where the tacit knowledge usually not conform to a standard knowledge model (i.e. produced and reproduced, some workers hesitate to free text). Therefore, we believe there is a need for share their knowledge and this would adversely knowledge models that provide decent influence the effectiveness of meetings as a result. characterisation of tacit knowledge and to integrate Becker [1] justifies this attitude by, “reluctance of both tacit and explicit knowledge fragments. The some members to speak, due to their shyness, low target models also have to cater for managing the

ISSN:1790-5109 133 ISBN: 978-960-474-002-4 2nd EUROPEAN COMPUTING CONFERENCE (ECC’08) Malta, September 11-13, 2008

uptodatedness of captured knowledge based on a repository. K-Assets represent the smallest level of defined lifecycle. These requirements we believe granularity of the knowledge repository. From the will have potential to make our model workable. perspective of the software engineering domain, any useful fragment of software development knowledge can be regarded as a K-Asset. For example, any 4 The proposed knowledge model lesson learned or process model or knowledge- A model is a purposeful abstraction of some part of embedded software artefact can be considered as K- reality [5]. It is meant to encapsulate the conceptual Asset. Individual knowledge workers can propose framework of the reality being modeled. Basically, what they consider as valuable software engineering our model is built around the notion of Knowledge K-Assets. These K-Assets are then scrutinized by Asset (K-Asset) as the basic building block of the peer knowledge workers subsequently. potential software-engineering knowledge

Process Model Software Competency M Used-in N

N Knowledge Asset Produces M N Proposes (i.e. seeds) M

Is linked N Has with

1 Argues 1 N Knowledge-embedded

software artefact N

Posts Confronts Argument N 1

Fig. 1 Conceptual view of the software-engineering Knowledge Model

Figure 1 shows the higher-level abstraction middleware between the knowledge workers and the of the proposed knowledge model. It shows the knowledge-embedded software artefacts is the stages of the lifecycle of software engineering process models that needs to be followed. For knowledge. Based on the proposed knowledge example, in the case of any C programmer (i.e. model, the knowledge generation loop starts from knowledge worker), S/he must have mastery level of characterizing the building blocks of the experience C programming (i.e. software competency). S/he knowledge. We believe the building blocks of the then can follow the structured programming concept software experience are the knowledge workers and (i.e. process model) in order to produce any piece of the set of software/hardware competencies they software code. Supposing that the produced software master and the knowledge-embedded software code is an error trapping routine since this reusable artefacts they produce. Any knowledge-embedded code has the potential to be embedded in any C software artefact is in fact generated using one or a based software. This code can then be characterized combination of software development skills in the as reusable knowledge fragment (i.e. knowledge- form of software/hardware competencies. As a embedded software artifact).

ISSN:1790-5109 134 ISBN: 978-960-474-002-4 2nd EUROPEAN COMPUTING CONFERENCE (ECC’08) Malta, September 11-13, 2008

Of course nothing can be characterized as a was contradicted by ibn al-Haytham (c. 965-1040), “knowledge fragment” simply based on the who is regarded to be the father of modern optics. assessment of individual knowledge workers. ibn al-Haytham correctly argued that humans see Sometimes, the characterization of what can be objects because the sun's rays of light are reflected considered as knowledge asset is a controversial from the objects into our eyes. This proves the fact issue because different knowledge workers might that knowledge validity is dynamic. Therefore, disagree about the validity of certain knowledge based on the proposed model, the process of assets. What might be proposed by a knowledge qualifying the proposed knowledge fragments relies worker as a knowledge fragments might not be on the consensus of other knowledge experts. This qualified as such by others who might be more consensus is realized through other experts’ knowledgeable. Moreover, the changing of opinions (i.e. arguments) in relation to qualifying or circumstances and newly emerged facts might disqualifying the proposed K-Assets. These influence the context of previously characterized arguments includ qualifying/disqualifying criteria knowledge assets. This is the reality of the posted by peer knowledge workers make up the knowledge phenomena anyway. For example, at one context associated with the captured knowledge. point of time, Ptolemy’s theory (c. 2nd century CE) This collaborative knowledge filtering constitutes a that we see objects because of the rays of light pool for generating tacit knowledge in the form of emanating from the eyes was very acceptable. With views, suggestions, and all types of arguments (see time and greater understanding, Ptolemy's theory the shaded portion in figure 1).

Post_Objection_Argument() [SupportingArgument_statistics=0]

Increases relevance Lower the relevance weight of a K-Asset weight of a K-Asset Inactive K-Asset

Post_Supporting_Argument() Post_Objection_Argument()

More relevant K-Asset Active K-Asset Less relevant K-Asset

Discard (weak K-Asset)

Irrelevant K-Asset (dead)

Fig. 2 K-Asset’s UML State diagram

continuously examine what it knows and change its Since, knowledge may become useless business beliefs accordingly. when the situation changes. Therefore, any captured In the case of the proposed knowledge knowledge has to be continuously examined and the model, tacit knowledge generated through the relevance status of individual K-Assets changes collaborative knowledge filtering helps to maintain accordingly. This goes inline with Nonaka’s claim the up-datedness of the captured knowledge. Figure that a company is not a machine but a living 2 shows UML state chart diagram that demonstrates organisation [12]. We also believe that an how the captured tacit maintains the dynamism of organisation is living in the sense that it has to the relevance of captured K-Assets. Newly proposed

ISSN:1790-5109 135 ISBN: 978-960-474-002-4 2nd EUROPEAN COMPUTING CONFERENCE (ECC’08) Malta, September 11-13, 2008

K-Assets are considered inactive and only become LiSER which is a software-engineering knowledge relevant as it receives the agreement and support management system. It is meant to capture not only from other knowledge workers. Likewise, the more explicit knowledge but also tacit knowledge in the objections it receives the less relevant it becomes. form of frequent arguments posted by knowledge The model discussed has been implemented through workers (see Figure 3).

K-Asset main attributes Explicit knowledge available through the URL provided.

URL of external explicit knowledge

Tacit knowledge in the form of argumentation map

Fig. 3 An implementation of the proposed model in the form software-engineering knowledge management system

the context behind the creation and validation 5 Conclusion criteria of captured knowledge. In order to maintain Knowledge is the winning factor for today’s the validity of captured knowledge, a knowledge organisations and knowledge management systems lifecycle is defined where the captured knowledge is constitutes the tool through which organisations continuously examined by peer knowledge workers, know what they know and what they should know. and the relevance of respective knowledge assets are Unfortunately, knowledge management systems determined accordingly. were structured more on formally structured knowledge while less attention is paid to capturing References: tacit knowledge. Software engineering is a very [1] Becker, K. (1997). Integrating Voting knowledge-intensive domain and a great portion of Techniques into a Discussion Rationale Model- software organisations’ knowledge is realised in based GDSS. ISDSS'97, Retrieved June 26 2002, tacit form. It is usually held in professionals’ heads from: http://inforge.unil.ch/isdss97/papers/71.htm as practical know-how that might be lost at any [2] Berlin, L.M., et al. (1993). Where did you put it? time. Therefore, there is an urgent need to capitalise Issues in the design and use of a group memory. tacit knowledge as well. This paper proposes a Proceedings of the INTERCHI’93 Conference on knowledge model that caters for capturing both tacit Human Factors in Computer Systems, 23-30, New and explicit knowledge in the software-engineering York: ACM. domain. The model integrates both explicit [3] Cummings, N. (1999). How OR conserves knowledge in the form of software artefacts and tacit precious resources in the hire & fire age. knowledge in the form of arguments that constitute Operational Research Newsletter, January 1999.

ISSN:1790-5109 136 ISBN: 978-960-474-002-4 2nd EUROPEAN COMPUTING CONFERENCE (ECC’08) Malta, September 11-13, 2008

[4] Conklin, E. J. (1996). Designing Organisational [9] Lee, J. (1991). Extending the Potts and Bruns Memory: Preserving Intellectual Assets in a Model for Recording Design Rationale, IEEE. . Retrieved May 2000, from : [10] Liebowitz, J. (2001). Knowledge Management: http://www.gdss.com/DOM.htm Learning from Knowledge Engineering. pp57-62, [5] Schreiber, G., et al. (2000). Knowledge ISBN 0-8493-1024-5, CRC Press, USA. engineering and management: the CommonKADS [11] MacLean, A., Young, R.M., Bellotti, V., methodology. Cambridge, Mass. : MIT Press, Moran, T. (1991). Questions, Options, and Criteria: ISBN: 0262193000. Elements of Design Space Analysis. Human- [6] Johannessen, J. A., Olaisen, J., Olsen, B. (2001). Computer Interaction, 6, 3 & 4, pp. 201-250. Mismanagement of tacit knowledge: the importance [12] Nonaka, I. (1998). The Knowledge Creating of tacit knowledge, the danger of information Company, Harvard Business Review on Knowledge technology, and what to do about it, International Management, Harvard Business School Press. Journal of Information Management, 21, 3-20. [13] Rittle, H., Webber, M. (1973). Dilemmas in a [7] Kokkoniemi, J. (2006). A Preliminary Model for general theory of planning. policy science. Generating Experience Knowledge Based Artifacts. [14] Tiwana, A. (2000). Knowledge Management Proceedings of the 39th Annual Hawaii Toolkit, The: Practical Techniques for Building a International Conference on System Science Knowledge Management System, , Prentice Hall (HICSS'06, track 7, p.153b. PTR. [8] Koskinen K., Pihlanto P., Vanharanta H. (2003). [15] Zack, M. H. (1999). Managing codified Tacit knowledge acquisition and sharing in a project knowledge, Sloan Management Review 40 (4), p. work context. International Journal of Project 45-58. Management, Volume 21, Number 4, May 2003, pp. 281-290(10), Elsevier Science.

ISSN:1790-5109 137 ISBN: 978-960-474-002-4