(IJCNS) International Journal of Computer and Network Security, 121

Vol. 2, No. 10, 2010

A Modified Framework of Knowledge Management

System Components for Collaborative Software
Maintenance
Mohd Zali Mohd Nor1, Rusli Abdullah2, Masrah Azrifah Azmi Murad3 and Mohd Hasan Selamat4
1
Universiti Putra Malaysia, Faculty of Computer Science and Information Technology,
43400 Serdang, Malaysia
[email protected]
2
Universiti Putra Malaysia, Faculty of Computer Science and Information Technology,
43400 Serdang, Malaysia
[email protected]
3
Universiti Putra Malaysia, Faculty of Computer Science and Information Technology,
43400 Serdang, Malaysia
[email protected]
4
Universiti Putra Malaysia, Faculty of Computer Science and Information Technology,
43400 Serdang, Malaysia
[email protected]

organization. The interrelations between these levels not

Abstract: Knowledge Management System (KMS) is critical to
software maintenance (SM) due to highly complex, knowledge- only affects the way knowledge is shared and transferred,
driven and collaborative environment. We propose a KMS but may also becomes barrier to efficient knowledge flows
framework for collaborative SM environment to model the [36].
requirements for sharing and sustaining knowledge in SM However, there are various issues associated with the SM
environment. The initial framework was based on literatures on knowledge, which makes organizing, storing, sharing and
Knowledge Management (KM), KMS and SM frameworks to
disseminating knowledge difficult. Among the problems are:
identify the knowledge, contents and technology components
related to SM environment. To verify the model, questionnaires
• Some of these skills and expertise are documented as
are subjected to a pilot study before being sent out to users and explicit knowledge, but more are hidden as tacit
maintainers. The final questionnaire survey responses were knowledge due scarcity of documentation [33].
analyzed using Rasch methodology. As a result, several less • Maintainers have to collaborate with colleagues and
important components were excluded from the initial model. other parties to obtain various information to enable
The revised model shall be further used in our ongoing study to them to carry out their software maintenance tasks.
develop a tool to assist SM Community of Practice (CoP)
• Domain knowledge are becoming more important to
members to perform their activities better.
software maintainers but are seldom stored in KMS or
Keywords: Knowledge Management, Knowledge Management other electronic means [25]. Often, maintainers have to
System, Software Maintenance, Collaborative Environment
rely on experts and also codes to understand the details
[9][42]. While domain knowledge are important to
1. Introduction maintainers, they are lacking, not stored properly or
Within software engineering activity cycle, software not readily available [33]. As such, maintainers,
maintenance (SM) has yet to receive proper attention [29] It especially newcomers, spend a lot of time searching,
is a costly process, where previous works [14][23][29][34] collaborating and understanding these knowledge.
estimated SM costs of between 60% to 90% of total software Changes to domain knowledge due to enhancements or
life cycle costs. The motivation in applying KM in SM in new business practices often affect usage of software
driven by the fact that the activities are knowledge-intensive application.
and depend largely on expertise of the maintainers, • Many SM tools are still not integrated to allow
However, SM organizations often have problems identifying seamless information combination, which hampers
resources and use of knowledge [30]. Managing knowledge information acquisition and sharing [25].
in this area is therefore critical to ensure that maintainers To address the above issues of knowledge within SM
can perform SM activities properly and timely, by sharing environment, a KMS framework shall be proposed. KMS
and obtaining vital knowledge. framework is required to ascertain that KM requirements
SM organization consists of individuals (users and are fulfilled, includes the necessary conceptual levels and
maintainers) working in interconnected groups/teams called support integration of the individual, team and
Community of Practices (CoP). Therefore, knowledge flows organizational perspectives [38].
not only within individual, but also within teams and Before the KMS framework in SM environment is proposed,
concepts of KM, KMS and SM activities shall be reviewed,
122 (IJCNS) International Journal of Computer and Network Security,
Vol. 2, No. 10, 2010

as follows: Meanwhile, KM frameworks for organization structure

includes Szulanski’s model of knowledge transfer [36],
Knowledge, KM and KMS Frameworks
APQC’s organizational KM model [4], Choo’s model of
As an overview, knowledge is defined as “a fluid mix of knowing organization [8], Selamat et al.’s KM framework
framed experience, values, contextual information and with feedback loop [35] and Holsapple and Joshi’s 3-fold
expert insight that provides a framework for evaluating and collaborative KM framework [19]. This framework
incorporating new experiences and information. It synthesizes the knowledge resources from Leonard-Barton,
originates and is applied in the mind of knowers. In and Petrach and Sveiby models; KM activities from Nonaka,
organizations, it often becomes embedded not only in APQC, Wiig, Van der Spek and Alavi’s models, and KM
documents an repositories but also in organizational influences from Wiig, APQC, Van der Speck, Szulanski and
routines, processes, practices and norms.” [10]. Leonard-Barton models..
Meanwhile, KM, in technical perspective, is defined as To transform KM framework into a system, KMS
the strategies and processes of identifying, understanding, framework is conceptualized. KMS is defined as “I.T-based
capturing, sharing, and leveraging knowledge system developed to support and augment the organizational
[2][3][10][35]. One of the major challenges is to facilitate process of knowledge creation, storage, retrieval, transfer
flow of knowledge not only within individual, but also and application” [3]. In general, a KMS framework consists
within teams and organizations [3]. of influential factors of KMS initiatives and their
For individual knowledge cycle, many KM frameworks interdependent relationships and a model of KMS
have been formulated [27][41], which were based on implementation [15]. However, systems and technology
Polanyi’s tacit and explicit knowledge. Nonaka and alone does not create knowledge [10], various other social
Takeuchi’s models the knowledge creation components as “incentives” and organizational strategy and culture are
socialization, internalization, combination and often required to stimulate use of technology to share
externalization. This SECI model has been used and knowledge. As an example, the KMS frameworks by Meso
synthesized by many others to model the KM for team and and Smith and Abdullah et al. are illustrated in Figure 2 and
organization levels. In SM, usage of SECI model can be Figure 3, respectively.
depicted in per Figure 1.
TECH NOLO G IES FUNCT IONS

Tacit To Tacit Knowledge Tacit To Explicit Knowledge Com puter-Media ted

Collaboration
Via Socialization Via Externalization
El ectronic task
SM knowledge are exchanged Articulate tacit knowledge into O RG ANIZATION AL
KM
managem ent
KNOW LEDGE

through experience sharing, explicit via concepts, SYSTEM Messaging

USI NG
K NOW LEDG E Know-how
brainstorming, observation metaphor, or models. In SM Video Confere nc ing Know-wha t
and practice. cases, these could be in form of & Visualiz at ion
Know-why

Today technologies: screenshots of errors, shadow Group Decision

Support
Se lf-M oti vat ed
Crea tivity
sessions, emails, conversations
Collaboration tools - W eb Browsing Person al T ac it

teleconferencing, desktop Today technologies: Data Mining

FIND ING Cultur al T ac it
K NOW LEDG E
Orga ni zational
video conferencing tools, Email, terminal sessions, chat Se arch & Re tri ev al
Ta cit

live-meetings, village wells, CREATING

K NOW LEDG E
Regu latory
Assets
Intel ligent Age nts
synchronous collaboration
Doc ument PACK AG ING
Explicit To Tacit Knowledge Explicit To Explicit Management
K NOW LEDG E

Via Internalization Knowledge Via Combination

Figure 8. Meso and Smith KMS Framework [24]
Knowledge is documented or Knowledge are combined,
verbalized, to help sorted, added , exchanged and
maintainers internalize and categorized, via specifications,
transfer knowledge, and also SCM entries and error analysis
help other maintainers to Today’s technologies:
‘reexperience’ bug scenarios.
Collaboration tools - E-mail,
Today technologies: GroupWare, Homepages,
Helpdesk and SCM consolidates in SCM. Data
applications are used to store mining to sort, and filter
bug reports and changes. information.
Visualization tool to read or
listen to success stories.
Figure 7. Nonaka's SECI Model (elaborated for SM)

When individuals collaborate and share knowledge, they

commonly do so within a team or group. Wenger defines
this group as the Community of Practice (CoP) - “a group of
people who share a concern, set of problem, or a passion
about a topic, and who deepen their knowledge and Figure 9. Abdullah et al. KMS Framework
expertise in this area by interacting on regular basis”. [40]
 (IJCNS) International Journal of Computer and Network Security, 123
Vol. 2, No. 10, 2010

Software Maintenance Knowledge Components for Collaborative SM.

Software maintenance (SM) is defined as “The totality of
activities required to provide cost-effective support to 2. Related Works
software system. Activities are performed during the pre- KMS for SM has been studied since late 1980s by Jarke
delivery stage as well as the post-delivery stage” [20]. and Rose [22], who introduced a prototype KMS to control
After studying the SM process flow in a SM organization database software development and maintenance, mainly to
[25], we envisage that the knowledge flow in SM facilitate program comprehension. The KMS is a decision-
environment is somewhat analogous to a river. Starting based approach that facilitates communication across time
from several streams or sources (call tickets, bugs reports, and among multiple maintainers and users, thus improving
testing bugs, request for enhancements, etc.) , the inputs maintenance support. Even though this research was carried
merge into the maintenance request mainstreams which go out quite some time ago, it provides the foundation on how
through many plains before ending at a delta of changed decisions could be assisted via shared knowledge. However,
objects, released applications and changes to domain rules. facilitating program comprehension is not enough as SM is
To support these flow of maintenance information, several more than just understanding codes and extracting
SM governance activities and tools are required. Tools such knowledge from codes.
as Helpdesk, Software Configuration Management (SCM), Similarly, Deraman introduced a KMS model for SM
Source Control, Project Management and others allow the which, albeit very simple, could provide us with the main
team and organization to monitor and control the processes. essence of SM knowledge – the Software Knowledge,
In addition, collaborative tools and platform allows users Change Request Knowledge and their functional interaction
and maintainers to communicate, cooperate and coordinate [11]. However, these alone, are not enough for users and
the required information to ensure good maintenance maintainers. Newer technologies such as software agents are
process. used to capture SM process knowledge in several researches
The knowledge required in SM can be summarized as [31][39]. However, no proper KMS framework was
follows [16][31][32]: conceptualized by these studies.
• Organizational knowledge, such as roles and resources. Looking at the wider perspective of software engineering
The parties involved in software maintenance activities (SE), KMS in SE have been studied by many, including
consist of various application users and software Santos et al. [33], Rus and Lindval [32] and Aurum et al.
maintainers. The list may include enduser, superuser, [5]. Rus and Lindval described the three main tasks of SE
maintenance manager, business analyst, systems (individual, team and organization) and identified the three
analyst, project manager, QA personnel, build level of KM support for each task. The 1st level includes the
manager, implementation personnel and trainer. core support for SE activities, document management and
Attached to these roles are the areas of expertise. competence management. Meanwhile, the 2nd level
• Managerial knowledge - such as resource management, incorporates methods to store organizational memory using
task and project tracking and management. method such as design rationale and tools such as source
• Technical knowledge – such as requirement analysis, control and SCM. The 3rd KM support level includes
system analysis, development tools, testing and packaged knowledge to support knowledge definition,
implementation. Critical to this is also the knowledge acquisition and organization. Although the illustrative
on supporting groupware and CASE tools such as model is not given, the above should describe the KMS
SCM, helpdesk and testing tools framework for SE. However, this model does not consider
• Domain knowledge – knowledge of the products and the social, physiological and cultural aspects of KM, as
business processes. identified by the previous other generic KMS frameworks.
• Knowledge on source of knowledge – where the Therefore, the main motivation for this study is to
knowledge resides, such as source codes, formulate a more detailed KMS framework for Collaborative
documentation, supporting CASE tools and more SM environment. The long-term goal of this study is to
importantly, the where the experts are. formulate a tool to support and automate KM tasks within
Based on past researches on KMS in SM, many collaborative SM environment. As such, the KMS
researchers concentrate only on SM processes, tools, framework shall place more emphasize on the technological
ontology, and knowledge flow. A proper study on KMS perspective.
framework in collaborative SM environment is still lacking.
The purpose of this paper is therefore to conceptualize such
3. Methodology
KMS framework.
This paper is structured as follows: section II discusses To formulate the KMS framework for Collaborative SM,
the related work on the related KMS frameworks and the following steps are used:
Section III reviews the methodology for the study. Then, 1. Review the existing KM and similar KM frameworks.
section IV discusses the proposed KMS component model 2. Review the existing KMS frameworks within the
for collaborative SM environment. Section V elaborates the related areas of SM.
findings of a questionnaire survey on the above KMS 3. List the KMS dimensions and components which
components and proposes a revised framework of KMS include the SM processes, tools and KM activities.
124 (IJCNS) International Journal of Computer and Network Security,
Vol. 2, No. 10, 2010

4. To verify the above dimensions and components, Rodriguez et al. [30]

suitable questionnaires are developed. • KM Activities are derived from Nonaka and Takeuchi
5. To ensure reasonable face validity, questionnaires were [27] and Holsapple & Joshi [19]. This includes
further deliberated and refined by academic lecturers in Acquiring knowledge, Selecting knowledge, using
Software Engineering, a statistician and several SM knowledge, Providing/ Creating knowledge and
managers. A pilot study of 13 respondents were Storing knowledge.
carried out to do the following: • SM governance tools are from Rus and Lindval [31],
a. Verify the constructs of the questions, based IEEE 14764 [20] and Mohd Nor et al. [25]. To support
on the responses given by a pilot group of these flow of SM information, tools such as Helpdesk,
respondents. SCM, Source Control and Project Management (PM)
b. Determine if the questions are well are crucial to monitor MRs.
understood by respondents (i.e. hard to • KM Components and Infrastructure are derived from
answer questions) Abdullah et al. [1], Meso & Smith [24], Dingsoyr &
c. Determine if the questions are mundane and Conradi [13], and Rus and Lindval [32] frameworks.
trivial (i.e. too easy) that perhaps the The major components includes computer-mediated
questions could just be left out. collaboration, Experience Mgmt System, Document
d. Rasch Unidimensional Measurement Model Management, KM portal, EDMS, OLAP, and
(Rasch) was used to analyze the above pilot Middleware tools.
data. Some problematic questionnaire items • Automation and knowledge discovery tools are from
are identified, revised and some discarded Meso and Smith [24], Abdullah et al. [1], Rodriguez et
to make the questionnaire more acceptable. al. [31] and new internet tools in the market. Tools
6. Distribute questionnaires to three in-house software such as GDSS, Intelligent Agents, Data
maintenance organizations in Klang Valley, Malaysia. mining/warehouse, Expert system and Case-Based
7. Analyze questionnaire responses using Rasch Reasoning (CBR). Active tools such as RSS are also
methodology to: useful to get the right knowledge to the right users at
a. Determine reliability of respondents and items the right time.
b. Determine outliers for both respondents and • KM Influences are derived from Holsapple and Joshi
items. [19] and Abdullah [2]. Among these are the
c. Determine the component groups’ cut-off points managerial influences and strategy, and psychological
to exclude the components from the model. This and cultural influences.
should give us the final KMS component model The questions are, what do users and maintainers want
for collaborative SM environment. from such a KMS, which of these components affect the
8. Revise the model for KMS Components for KMS and, most important of all, how does these
Collaborative SM Framework components affect KM activities in Collaborative SM? To
The questionnaire are developed using a 4-Likert Scale answer the above questions, questionnaires are drafted and
order of importance, with 1 being less important and 4 distributed to selected users and maintainers to gauge their
denotes more important. Evaluation of questionnaire results perspective on these KM components, mainly to explain the
shall be performed using Rasch. It is a probabilistic model following issues:
that uses ‘logit’ as the measurement units, by transforming
the ordinal data and raw scores into a linear scale [7]. Being
linear, Rasch enables us to conduct more constructive 4.1 The Important And Required General KM And
analyses. SM Knowledge
Required knowledge, such as organizational knowledge,
managerial knowledge, technical knowledge, business
4. Initial KMS Dimensions and Components domain knowledge and knowledge on source of knowledge
To build the list of dimensions and components for a are important in SM processes. However, in most of the
KMS framework, a review of current KMS framework for cases, we suspect that the business domain knowledge are
generic KMS, and related SE/SM KMS are conducted. The important, but not stored for maintainers to access.
theoretical constructs for KM framework, KMS framework 4.2 How Important Are SM Governance Tools As
and SM knowledge dimensions are summarized in Enablers In Knowledge Activities
Appendix B, Appendix C and Appendix D, respectively. Based on IEEE 14764, SM processes include process
Based on the construct summary, components suitable for implementation, problem and modification analysis, review
SM KMS framework are identified as follows: and acceptance, development/modification, migration and
• Required knowledge, such as organizational retirement. Whilst KM components deals with converting
knowledge, managerial knowledge, technical tacit to explicit and then storing the knowledge, the SM
knowledge, business domain knowledge and governance tools are used to manage the flow of information
knowledge on source of knowledge, are derived from and knowledge relating to the SM daily activities and
Ghali [16, Rus and Lindval [31], Dias et al [12] and processes. Tools such as Helpdesk application, SCM, source
 (IJCNS) International Journal of Computer and Network Security, 125
Vol. 2, No. 10, 2010

control and project management are used throughout the discussed in Section 4 above, are analyzed in three parts;
SM processes and hence provides a good deal of input to the data reliability, fitness of respondent data and questionnaire
knowledge contents. The questions are – how well and items data and determination of component groups cut-off
important are these tools in the KM activities (acquiring, points.
selecting, using, providing/creating and storing knowledge).
5.2 Data Reliability
4.3 The Importance Of KMS Foundation
Components And Infrastructure Summary statistics for respondents (person) and items
KMS foundation components include, among a few, (questions) are depicted in Table 1 and Table 2,
computer-mediated collaboration, Experience Mgmt System, respectively. 41 respondents returned the survey
Document Management, KM portal, EDMS, OLAP, and questionnaire. Out of which, Rasch identified an extreme
Middleware tools and Knowledge Map. While collaboration score which will later be excluded from further analysis.
tools allow users to impart and share both tacit and explicit
knowledge synchronously and asynchronously, the other Table 1. Summary of 40 Measured (Non-Extreme) Persons
tools are useful to search and extract available explicit Raw Cou Mea Infit Outfit
information. Knowledge map refers to navigation aid to Score nt sure mnsq zstd mnsq zstd
both explicit and tacit knowledge, by illustrating the mean 133.8 42.8 0.49 1.02 -0.2 1.01 -0.2
knowledge flows within the organization [17]. In many s.d. 14.9 3.5 0.69 0.52 2.1 0.53 2
aspects, it involves establishing knowledge ontology, max. 167 45 2.64 3.14 6.4 3.37 6.7
mapping/linking the knowledge and validating the min. 86 30 -0.65 0.28 -4.5 0.28 -4.4
knowledge map. Real RMSE =.30 Adj.SD=.62 Separation =2.10 Person Reliability= .82
Model RMSE =.27Adj.SD=.64 Separation=2.35 Person Reliability= .85
4.4 How Important Are Different Automations And
S.E. Of Person Mean = .11Maximum Extreme Score: = 1 Persons
Automation Tools To The Overall Activities Of KMS?
Valid Responses: 95.0%
Automation speeds up and assists maintainers in their
Person RAW SCORE-TO-MEASURE CORRELATION = .51
daily activities. Technologies such as knowledge-map, CBR,
(approximate due to missing data)
expert system, agent technology and RSS are useful to assist CRONBACH ALPHA (KR-20) Person RAW SCORE RELIABILITY =
users and maintainers to get the right knowledge at the right .94 (approximate due to missing data)
time.
4.5 How Important Are Managerial Influences And Table 2. Summary of 45 Measured Items
Strategies To The KMS Activities And Processes? Raw Infit Outfit
Managerial influences such as leadership, coordination, Score Count Measure mnsq zstd mnsq zstd
control and measurement [19] may affect the general SM mean 118.9 38 0 1 0 1 0.1
KM activities and processes. Strategy deals with how KMS s.d. 16.6 3.3 0.62 0.12 0.6 0.15 0.7
is planned for use, whether through codification (storing the max. 150 40 1.16 1.29 1.5 1.4 1.9
explicit information), or personalization (storing the min. 88 29 -1.2 0.83 -1.3 0.74 -1.3
knowledge map). Real RMSE = .32 Adj.SD =.54 Separation =1.69 Item Reliability = .74
Model RMSE = .27 Adj.SD = .64 Separation =2.35 Item Reliability =.75
4.6 How Important Are Psychological And Cultural
S.E. Of Person Mean = .09
Influences In The Overall Activities Of KMS?
Psychological issues include motivation, reward and
awareness. Meanwhile cultural influences include the trusts, The spread of person responses is = 3.29 logit is fair. This
beliefs, values, norms and unwritten rules.. is due to extreme responses by a person (code=PAUS2).
However, Reliability = 0.82 and Cronbach Alpha=0.94
indicates high reliable data and hence the data could be used
5. Discussion
for further analyses.
The Pilot study [26] revealed the item reliability in the On the questionnaire items aspects, the summary of 45
initial questionnaire was found to be poor and a few measured questionnaire items (see Table 3) reveals that the
respondents and items were identified as misfits with spread of data at 2.36 logit and reliability of 0.74 are good
distorted measurements. Some problematic questions are and fair, respectively. Details on measured items are listed
revised and some predictably easy questions are excluded in Appendix A. None of the items are beyond the critical
from the final questionnaire. measures (0.4 < Acceptable Point Measure Correlation < 0.8
In the final questionnaire survey, 41 respondents from and 0.5 < Outfit Mean Square < 1.5, and -2.0 < Outfit z-
three organizations participated in the survey. Among these, standardized value < 2.0). The previous pilot study is
27% are users and superusers, 22% are systems analysts, therefore proven helpful in making the questionnaire more
15% are programmers, 15% SM managers, 10% business reliable.
analysts and the rest are DBAs, Helpdesk and other
technical staff. In years of service, 40% of respondents have
more than 10 years experience, 34% have between 1 to 5
years, 24% have between 6 to 10 years and only 3% have
less than 1 year of service.
The results of the survey, based on the components
126 (IJCNS) International Journal of Computer and Network Security,
Vol. 2, No. 10, 2010

Table 3. Items Statistics – Measure Order

Mea-
Item sure Model Outfit
No. (Logit) mnsq Infit ZStd mnsq ZStd PT Mea Corr.
1 -1.18 0.98 0 1 0.1 0.31 B1 Roles
2 0.12 0.88 -0.7 0.84 -0.8 0.44 B2 Resources
3 -0.32 0.88 -0.6 0.85 -0.7 0.43 B3 SM_Tasks
4 -0.26 0.84 -0.8 0.81 -1 0.46 B4 Req_analysis
5 -0.4 0.97 0 1.13 0.6 0.37 B5 Sys_analysis
6 -0.29 0.92 -0.3 0.9 -0.4 0.43 B6 Dev_skill
7 -0.1 0.87 -0.5 0.88 -0.4 0.49 B7 Testing_skill
8 -0.01 0.84 -0.7 0.84 -0.8 0.48 B8 Implementation
9 -0.28 1.03 0.2 1.02 0.2 0.31 B9 Info_Source
10 -0.82 1.04 0.3 1 0.1 0.28 BA Domain
11 -0.56 1.06 0.3 1.1 0.4 0.24 CA SCM
12 -0.53 1.19 0.6 1.27 0.9 0.13 CB Helpdesk
13 0.05 0.9 -0.2 0.93 -0.1 0.48 CC VCS
14 0.11 0.83 -0.5 0.87 -0.4 0.56 CD PM
15 -0.72 0.89 -0.6 0.8 -0.7 0.31 D1 Email
16 0.92 1.27 1.2 1.25 1.1 0.38 D2 Fax
17 0.99 1.06 0.4 1.07 0.4 0.49 D3 Memo
18 0.15 1.17 0.6 1.17 0.6 0.28 D4 E-Group Figure 10. Person-Item Distribution Map
19 0.09 1.03 0.2 1 0.1 0.43 D5 BB
20 -0.36 1.06 0.4 1.04 0.3 0.3 D6 F2F
21 -0.36 0.97 -0.1 0.96 -0.1 0.36 D7 Phone Person PAUS2 (a User) being the highest in PIDM, have
22 1.06 1.29 1.5 1.35 1.7 0.29 D8 Chat the tendency to give high importance ratings to most of the
23 0.85 1.11 0.6 1.1 0.6 0.37 D9 Conference
24 0.83 0.86 -0.6 0.86 -0.6 0.53 DA Audio/Video
questionnaire items, whilst P4SA2 (a Systems Analyst)
25 -0.12 1.11 0.5 1.09 0.4 0.3 DB Portal tends to rate lower. On detailed post-questionnaire
26 0.47 1.25 1.2 1.24 1.2 0.24 DC Intranet inspection, person PAUS2 is a new user which is not
27 -0.45 1.06 0.3 1.05 0.3 0.32 DD Search
28 -1.2 0.99 0 1.03 0.2 0.29 DE Groupware
familiar with the technology components being asked in the
29 0.17 0.89 -0.2 0.88 -0.2 0.47 EA GDSS questionnaire and has tendency to give equal high ratings to
30 0.76 0.95 0 0.91 -0.1 0.31 EB CBR all question. Likewise, person P4SA2 is a long-serving user
31 0.2 0.83 -0.2 0.74 -0.3 0.56 EC Agent
who is familiar with the components. Since PAUS2 has
32 0.02 1.1 0.4 1.21 0.8 0.28 ED SMS
33 0.53 0.91 -0.2 0.89 -0.3 0.44 EE RSS been identified by Rasch as extreme, the person shall be
34 -0.09 0.86 -1.3 0.79 -1.3 0.39 F1 Leadership excluded from further analysis. This is reasonable, since our
35 0.81 0.9 -0.9 0.87 -1 0.41 F2 Coordination
priority is to evaluate the items and less on persons.
36 -0.25 0.9 -0.3 0.87 -0.4 0.49 F3 Control
37 0.75 0.95 -0.2 0.96 -0.2 0.45 F4 Audit On the Item side, the distribution is quite closely bunched
38 1.13 1.1 0.8 1.14 0.9 0.26 F5 Codification together, with no obvious outliers. Among these items, B1
39 -0.5 1 0.1 1.05 0.3 0.4 F6 Personalization Roles, BA Domain, DE Groupware, G1 Management and
40 1.16 1 0 1.01 0.1 0.35 F7 Combination
41 -0.99 1.16 0.9 1.4 1.9 0.2 G1 Mgmt G5 CoP are below the minimum measure of Persons. This
42 -0.74 0.97 -0.1 1.01 0.1 0.41 G2 Org_value indicates overall agreeableness on the high importance of
43 -0.08 1.02 0.2 1.04 0.3 0.35 G3 these components.
Hoard_knowledge
44 0.24 1.03 0.2 1.02 0.2 0.37 G4 Rewards 5.3 Component Group Cut-off Points
45 -0.79 0.98 0 0.93 -0.2 0.39 G5 CoP
avg 0 1 0 1 0.1 63.5 There are no hard and fast rules on how to determine
S.D. 0.62 0.12 0.6 0.15 0.7 11 which of the less important KMS components should be
excluded from the framework. We listed the components
and sorted them in descending logit values. The list was
Fitness of Respondent and Questionnaire Items data distributed to four experts from knowledge management and
software engineering fields, and three software maintenance
managers, to identify the cutoff point for important
components.
Rasch provides the Person Item Distribution Map Based on the overall experts’ judgments, the following
(PIDM), which is similar to histogram (see Figure 5). PIDM components are selected to be excluded from the model:
allows both person and items to be mapped side-by side on • B8 - Knowledge on implementation is important in my
the same logit scale to give us a better perspective on the line of work.
relationship of person responses to the items. PIDM • B2 - Knowledge on how to manage resources in SM
indicates a higher Person Mean (0.48) compared to the organization.
constrained Item Mean. This indicates tendency to rate • CD - Project Management tool to store and share
higher importance to the prescribed questionnaire items. explicit knowledge on SM resources and activities
• D2 - Electronic fax
• D3 - Paper memo
• D8 - On-line chat (MSN Messenger, Yahoo Chat, etc)
• EE - RSS technology disseminate information
 (IJCNS) International Journal of Computer and Network Security, 127
Vol. 2, No. 10, 2010

• EB - Expert Systems or CBR tools to finalize the important components for KMS framework
• F5 - Codification strategy (storing explicit information) for collaborative SM environment. This new framework
• F7 - Combination of both codification and shall be used in our ongoing study to develop a tool to assist
personalization strategies SM CoP members to perform their SM activities better.
• G4 – Rewards should be implemented to promote more
knowledge sharing Acknowledgment
This research is funded by the Malaysia Ministry of
6. Revised KMS Framework Science and Technology (MOSTI) e-ScienceFund Project
Based on the above reduced components, the revised No.01-01-04-SF0966.
framework is depicted in Figure 5, which consists of the
components evaluated using the questionnaire survey, and References
the following fixed components:
[1] R. Abdullah, Knowledge Management System in a
• SM Process/Activities are standard activities described Collaborative Environment, University Putra Malaysia
by ISO/IEC 14764 standard Press, 2008.
• The KM components and infrastructure are derived [2] R. Abdullah, S. Sahibuddin, R. Alias, and M.H.
from the other standard KMS infrastructure and Selamat, Knowledge Management System Architecture
frameworks. For Organizational Learning With Collaborative
Process/ Environment, International Journal of Computer
CoP Knowledge Tools KM Infra
Activity Science and Network Security, vol. 6, 2006.
Knowledge SM Activities [3] M. Alavi and D. Leidner, Knowledge Management
SM Tools
Users Required Process Helpdesk Systems: Issues, Challenges, and Benefits,
Org. Knowledge Implementation SCM
Superuser Technical Analysis Source Control KM Components Communication of AIS, vol. 1, 2000.
Knowledge Planning & & Infrastructure
Helpdesk Managerial Acceptance KM Portal [4] Arthur Anderson and APQC, The Knowledge
Knowledge Modification & Document
SM Manager Business Implementation Management Management Assessment Tool: External
Domain Migration & Automation & K-
Discovery Tools Directory
Business Analyst Knowledge Refinement services Benchmarking Version, Arthur Anderson/APQC,
Multi Agent
System (MAS) Knowledge Map
Systems Analyst
KM Activities Search Engine Middlewares 1996.
Data Mining & OLAP
Programmer Acquire Select Warehousing [5] Aurum, R. Jeffery, C. Wohlin, and M. Handzic,
knowledge knowledge Expert Systems
DBA Managing Software Engineering Knowledge, Springer,
Store
Others knowledge 2003.
Collaborative Components
Provide/Create Use Same Time Diff Space [6] G. Avram, Knowledge Work Practices in Global
knowledge knowledge Same Time & Space Phone . Email
Face-to-face Software Development, The Electronic Journal of
Groupware Audio Video
KM Soft influences Conference
Cultural influences Knowledge Management, vol. 5, 2007.
Psychological Trusts, Beliefs, Values, Diff Time,Same Space Diff Time, Diff Space
Motivation, & Awareness Norms & Unwritten Rules, Email . E-Group Portal . BB
Managerial influences K-Hoarding, CoP BB . Portal E-Group
[7] T. Bond and C. Fox, Applying the Rasch Model:
Leadership, Coordination, Strategy - Personalization Intranet Intranet
Fundamental Measurement in Human Sciences., New
Jersey: Lawrence Album Associates, 2007.
[8] C. Choo, An Integrated Information Model of the
Figure 11. The Revised Model for KMS Components for
Organization: The Knowing Organization, 1996.
Collaborative SM [9] S. Das, W. Lutters, and C. Seaman, Understanding
Documentation Value in Software Maintenance,
7. Conclusion and Future Works Proceedings of the 2007 Symposium on Computer
Symposium On Computer Human Interaction For The
In SM environment, KMS are critical to ensure that KM Management Of Information Technology, 2007.
activities such as knowledge acquisition, storage and [10] T. Davenport and L. Prusak, Working Knowledge:
retrieval and processes includes not only the hard- How Organization Manage What They Know, Harvard
components (tools and infrastructure), but also the soft- Business School Press, 2000.
components (managerial, psychological and cultural). [11] Deraman, A Framework For Software Maintenance
To formulate the KMS framework for collaborative SM, Model Development, Malaysian Journal of Computer
Science, vol. 11, 1998.
the components on KMS, SM governance, and automation
[12] M. Dias, N. Anquetil, and K. Oliveira, Organizing the
and knowledge discovery are compiled from various Knowledge Used in Software Maintenance, Journal of
literatures. An initial model of modified KMS components Universal Computer Science, vol. 9, 2003.
for collaborative SM is proposed. The relationships between [13] T. Dingsoyr and R. Conradi, A Survey Of Case Studies
these components are used to construct the questionnaire, Of The Use Of Knowledge Management In Software
which were first tested in a pilot study. A pilot study was Engineering, International Journal of Software
undertaken to evaluate the construct validity of the Engineering and Knowledge Engineering, vol. 12,
2002.
questionnaire, as well as identifying the expected measures. [14] R. Fjeldstad and W. Hamlen, Application Program
A survey using revised questionnaire items was carried Maintenance Study: Report to Our Respondents,
out in three SM organizations in Klang Valley Malaysia to Software Engineering- Theory and Practices, Prentice
gain a better perspective on the importance of the SM and Hall, 1998.
KM components. As a result, several less important [15] S. Foo, A. Chua, and R. Sharma, Knowledge
components were excluded from the initial model. The management Tools and Techniques, Singapore:
Pearson Prentice Hall, 2006.
revised model was further deliberated, by experts’ opinion,
128 (IJCNS) International Journal of Computer and Network Security,
Vol. 2, No. 10, 2010
[16] N. Ghali, Managing Software Development Categories: Survey versus Measurement, Journal of
Knowledge: A Conceptually-Oriented Software Empirical Software Engineering, vol. 8, 2003.
Engineering Environment, PhD. Thesis, University of [35] M. Selamat, R. Abdullah, and C. Paul, Knowledge
Ottawa, Canada, 1993. Management System Architecture For Organizational
[17] D. Grey, Knowledge Mapping: A Practical Overview. Learning With Collaborative Environment,
Available at: http://kmguru.tblog.com/post/98920, International Journal of Computer Science and
Accessed September 30, 2009 Network Security, vol. 6, 2006.
[18] M. Handzic and H. Hasan, The Search for an [36] K. Sherif, Barriers to Adoption of Organizational
Integrated KM Framework, Australian Studies in Memories: Lessons from Industry, Knowledge
Knowledge Management, UOW Press, 2003. Management Systems: Theory and Practice, Thomson
[19] C. Hosapple and K. Joshi, Understanding KM Learning, 2002
Solutions: The Evolution of Frameworks in Theory and [37] G. Szulanski, Exploring Internal Stickiness:
Practice, Knowledge Management Systems: Theory Impediments to the Transfer of Best Practice Within
and Practice, Thomson Learning, 2002. The Firm, Strategic Management Journal, vol. 17,
[20] IEEE, IEEE 14764-2006, Standard for Software 1996.
Engineering - Software Life Cycle Process - [38] F. Ulrich, A Multi-Layer Architecture for Knowledge
Maintenance, The Institute of Electrical and Management Systems, Knowledge Management
Electronics Engineers, Inc., 2006. Systems: Theory and Practice, Thomson Learning,
[21] V. Janev and S. Vranes, The Role of Knowledge 2002, pp. 97-111.
Management Solutions in Enterprise Business [39] Vizcaíno, J. Soto, and M. Piattini, Supporting
Processes, Journal of Universal Computer Science, Knowledge Reuse During the Software Maintenance
vol. 11, 2005. Process through Agents, Proceedings of the 6th
[22] M. Jarke and T. Rose, Managing Knowledge about International Conference on Enterprise Information
Information System Evolution, Proceedings of the Systems (ICEIS), 2004.
1988 ACM SIGMOD International Conference, 1988. [40] E. Wenger, Communities of Practice: Learning,
[23] B. Lientz and E. Swanson, Characteristics of Meaning, and Identity, Cambridge University Press,
Application Software Maintenance, Communications 1998.
of the ACM, vol. 24, 1981. [41] K. Wiig, Knowledge Management Foundation, Schema
[24] P. Meso and R. Smith, A Resource-Based View Of Press, 1993.
Organizational Knowledge Management Systems, [42] T. Wilson, The nonsense of knowledge management,
Journal of Knowledge Management, vol. 4, 2000. Journal of Information Research, Vol. 8, 2002.
[25] M.Z. Mohd Nor and R. Abdullah, A Technical
Perspective of Knowledge Management in Mohd Zali Mohd Nor received the B.Sc. in
Collaborative Software Maintenance Environment, Mathematics from the University of
Knowledge Management International Conference Michigan, Ann Arbor in 1988 and Master of
(KMICE), 2008. Management in I.T. from Universiti Putra
[26] M.Z. Mohd Nor, R. Abdullah, M.A. Azmi Murad, M. Malaysia in 2005. He is now an I.T
H. Selamat, A. A. Aziz, "KMS Components For Development and Maintenance manager in a
Collaborative Software Maintenance – A Pilot Study", shipping company, whilst pursuing his Ph.D
International Conference on Information Retrieval and in University Putra Malaysia. His main area
Knowledge Management (CAMP10), 2010. of interest is Knowledge Management in Collaborative Software
[27] G. Natarajan and S. Shekar, Knowledge Management: Maintenance.
Enable Business Growth, McGraw-Hill, 2001.
[28] Nonaka and H. Takeuchi, The Knowledge-Creating
Company, New York: Oxford University Press, Inc.,
1995.
[29] T. Pigoski, Software Maintenance, Guide to the
Software Engineering Body of Knowledge
(SWEBOK), The Institute of Electrical and Electronics
Engineers, Inc., 2004.
[30] O. Rodriquez, A. Martinez, J. Favela, A. Viscaino, and
M. Piattini, Understanding and Supporting Knowledge
Flows in a Community of Software Developers,
Lecture Notes in Computer Science, vol. 2198, 2004.
[31] O. Rodriquez, A. Martinez, J. Favela, A. Viscaino, and
M. Piattini, How to Manage Knowledge in the
Software Maintenance Process, Lecture Notes in
Computer Science, vol. 3096, 2004.
[32] Rus and M. Lindvall, Knowledge Management in
Software Engineering, IEEE Software, vol. 19, 2001.
[33] G. Santos, K. Vilela, M. Montoni, and A. Rocha,
Knowledge Management in a Software Development
Environment to Suport Software Process Deployment,
Lecture Notes in Computer Science, vol. 3782, 2005.
[34] S. Schach, B. Jin, L. Yu, G. Heller, and J. Offutt,
Determining the Distribution of Maintenance