KMDL - Capturing, Analysing and Improving Knowledge-Intensive
(University of Potsdam, Germany
(University of Potsdam, Germany
(University of Potsdam, Germany
Abstract: Existing approaches in the area of knowledge-intensive
processes focus on integrated knowledge and process management systems,
the support of processes with KM systems, or the analysis of knowledge-intensive
activities. For capturing knowledge-intensive business processes well known
and established methods do not meet the requirements of a comprehensive
and integrated approach of process-oriented knowledge management. These
approaches are not able to visualise the decisions, actions and measures
which are causing the sequence of the processes in an adequate manner.
Parallel to conventional processes knowledge-intensive processes exist.
These processes are based on conversions of knowledge within these processes.
To fill these gaps in modelling knowledge-intensive business processes
the Knowledge Modelling and Description Language (KMDL) got developed.
The KMDL is able to represent the development, use, offer and demand of
knowledge along business processes. Further it is possible to show the
existing knowledge conversions which take place additionally to the normal
business processes. The KMDL can be used to formalise knowledge-intensive
processes with a focus on certain knowledge-specific characteristics and
to identify process improvements in these processes. The KMDL modelling
tool K-Modeler is introduced for a computer-aided modelling and analysing.
The technical framework and the most important functionalities to support
the analysis of the captured processes are introduced in the following
Keywords: process-oriented knowledge management, knowledge-intensive
business processes, knowledge modeling description language, K-Modeler
There are two main approaches to knowledge management distinguished
in the literature [Mentzas, 03]. The process-centred approach treats KM
as an interpersonal communication process. The product-centred approach
on the other hand focuses on the artefacts for knowledge, i.e. the documents,
their creation and reuse in corporate computer-based systems.
In the last few years the process-oriented knowledge management as integration
of business process management and knowledge management has been established
in the scientific and practical field.
The process-oriented knowledge management not only considers the business
processes but uses the process-oriented view to describe the dynamic knowledge
conversions between the process participants. Knowledge and business processes
are connected directly; therefore the integrated consideration is indispensable.
Business processes can be modelled and analyzed extensively with well
known and established methods. Further approaches exist that consider knowledge
as a component of a company or an organization [Goesmann,
02], [Remus, 02b]. The simple mapping of static
knowledge (typically in an explicit manner as information) does not fulfil
the requirements of a comprehensive and integrated approach for process-oriented
knowledge management. Only the coordination of business processes with
the processes of knowledge processing guarantees an efficient knowledge
flow [Remus, 02b]. The above mentioned problems and
challenges have been the trigger for the development of the Knowledge Modelling
and Description Language KMDL and the tool K-Modeler to model and analyze
knowledge-intensive business processes.
1.1 Overview of the Contents
In this contribution, after introducing business process oriented knowledge
management knowledge-intensive business processes are defined. In the following
section the theoretical foundation of the KMDL, the differentiation of
tacit and explicit knowledge and the knowledge conversion are described.
These concepts are used to evaluate existing tools and methods for process-oriented
knowledge management to point out the need for a new method to capture,
model and analyse knowledge-intensive business processes.
The fourth section explains the KMDL object model. Each object is shortly
described and the obligatory and optional attributes are presented. A practical
example is used to create a better understanding of the modelling technique.
The fifth section gives an overview of the used KMDL procedural model.
It consists of six phases which are explained. The second phase is divided
into a sub-procedure to elicit correctly all required information.
The knowledge modelling tool K-Modeler and its main functionalities
are depicted in the sixth section. First of all the integration of the
tool into Eclipse is explained. Consecutively the modelling, the process
analysis, the defined views on the model, the support of skill management
applications and the XML data description are shortly described. The next
section illustrates the practical benefits of KMDL by applying the language
in practical projects. In one case the KMDL was used to improve the communication
between the product development and the customer care. The last paragraph
introduces the present and the future work of KMDL.
The following figure 1 shows the content of the contribution.
Figure 1: Overview of the Contents of this Contribution
2 Process-Oriented Knowledge Management
Abecker identifies a field of research, which utilizes the modelling
of business processes to enable the derivation of knowledge management
measures [Abecker, 02]. The three application scenarios
"business processes as initial point for knowledge management",
"knowledge management and process execution" and "business
processes as subject of knowledge management" require some or all
of the following project stages: Systems Design (consisting of Systems
Planning, Analysis, and Implementation), Systems Usage, and Systems Evolution.
To exactly classify different research approaches, it can be further segregated
into three different layers. On the top layer, strategic business process-oriented
knowledge management is a top-down perspective, which derives knowledge
objectives from the long-term business objectives. The bottom layer deals
with KM design based on communication analysis and diagnosis. It primarily
deals with communication aspects of knowledge work and develops appropriate
methods or tools. It is thus very hard to be separated from the middle
layer, where Abecker allocates approaches of business process-oriented
design, where methods and tools for business process analysis are extended
to meet the new requirements of knowledge management. This middle layer
is dealing with modelling methods derived from business process management
and the modelling of existing processes to find potentials for improvement.
A selection of the existing approaches BPO-KM, PROMOTE, and CommonKADS
which belong to this category are introduced further in [Trier,
04]. The following short introduction of the three approaches is based
on the result of this analysis.
The first selected approach is BPO-KM (in German: GPO-WM®). It proposes
a method for a process-oriented analysis and design of knowledge management
solutions [Heisig, 03]. Within this procedure of eight
steps, the KM audit analyses the fundamental conditions including the evaluation
of existing IT systems, the analysis of the information- and knowledge
culture, and the determination of the demand for information and knowledge.
The main focus of this step is the identification of potentials for
improvement of the existing utilisation of knowledge in the business context.
The subsequent step analyses knowledge-intensive processes to identify
strengths and weaknesses or possible improvements. Further the process-
and task-related demand for knowledge is identified.
Another approach is the PROMOTE method, which integrates strategic planning
with the evaluation of knowledge management and business process management
[Hinkelmann, 03]. The intended scope of the approach
covers the analysis, the modelling, and the execution of knowledge-intensive
processes. It extends the more general method of business process management
systems (BPMS) including strategic decision, reengineering and resource
allocation, and workflow and performance evaluation [Hinkelmann,
03]. The additional KM related steps are creating awareness for enterprise
knowledge, discover knowledge processes, create operational knowledge processes
and organisational memory, and evaluate enterprise knowledge. Next to the
process-oriented KM models introduced, the established knowledge engineering
approach CommonKADS could influence a method for the capturing of knowledge-intensive
business processes. Although its objective of constructing a program that
can perform a difficult task adequately is completely different [Schreiber,
00], its process of knowledge acquisition can be regarded as similar,
because knowledge acquisition includes the elicitation, collection, analysis,
modelling, and validation of knowledge for knowledge engineering and knowledge
management projects. The according knowledge acquisition (KA) techniques
have been developed to help with the elicitation of knowledge from an expert.
3 Knowledge-Intensive Business Processes
Within process-oriented knowledge management the knowledge-intensive
business process is the primary perspective [Remus, 02b].
Several attempts have been made in the literature to define knowledge-intensive
business processes. Heisig points out the opportunity to schedule the knowledge
demand and evaluates knowledge-intensity according to the existence of
variability and exceptions [Heisig, 02]. Other sources
define processes knowledge-intensive if an improvement with conventional
methods of business reengineering is not or only partially possible [Remus,
02a]. Davenport recognizes the knowledge-intensity by the diversity
and uncertainty of process input and output [Davenport,
95]. A process is knowledge-intensive if its value can only be created
through the fulfilment of the knowledge requirements of the process participants.
Several properties which are typical for knowledge-intensive business processes
are introduced in the following list:
- In knowledge-intensive processes, knowledge contributes significantly
to the values added within the process. Innovation and creativity play
a major role in such processes [Eppler, 99]. People
within the process have a large scope in the freedom of decision, they
can decide autonomously.
- The event flow of knowledge-intensive business processes is not clear
in advance, as it can evolve during the process [Davenport,
- The participants in the process have different experiences and bring
in knowledge from different domains at different levels of expertise [Heisig,
- The life-time of knowledge involved in the process is often very short
[Eppler, 99], it is outdated very fast. It is usually
very time-intensive to build up this knowledge [Schwarz,
- Knowledge-intensive business processes often do not follow structured
working rules and often lack metrics for evaluating the success of the
process [Davenport, 00].
- The IT-support for knowledge-intensive business processes is generally
not very sophisticated because it strongly relies on socialization and
informal exchange of knowledge [Hoffmann, 02].
- A knowledge-intensive process should be a core process of the company
and it should produce or add new knowledge to the organization's knowledge
base [Hamel, 90].
- Often the costs of knowledge-intensive processes are very high.
Looking at these criteria, we can classify various processes as knowledge-intensive.
Just two examples are software development processes [Kidd,
94] or processes in public administration.
Common business processes are characterized by a predefined process
structure and repeated tasks that are fulfilled basing on the underlying
process model, which contains information, tasks and user roles. Knowledge-intensive
business processes are only partially mapped by the process model due to
unpredictable decisions or tasks guided by creativity. Typically knowledge
flows and knowledge transfers between media and persons are necessary to
achieve a successful process completion.
Identifying, modelling, analyzing and finally optimizing knowledge-intensive
processes should be the long-term objective of a process-oriented knowledge
management approach [Gronau, 04c]. Knowledge management
and business processes are integrated and should be evaluated as a whole
4 Theoretical Foundation of KMDL
This section introduces the theoretical concepts which are used to define
the Knowledge Modeling and Description Language. The first paragraph outlines
the tacit and explicit knowledge defined by Nonaka and Takeuchi. In the
second paragraph the concept of knowledge conversion will be introduced.
These concepts were used to analyze and evaluate existing tools for modelling
and analyzing knowledge-intensive business processes. Finally the requirement
of a new language specification is pointed out.
4.1 Tacit and explicit knowledge
The fundamentals of the process-oriented knowledge modelling language
KMDL (Knowledge Modeling and Description Language) are influenced by the
ideas of Nonaka and Takeuchi [Nonaka, 95]. In their
book Nonaka and Takeuchi have build a whole theory about knowledge and
the creation of knowledge. This theory is based on the distinction between
tacit and explicit knowledge.
The term tacit knowledge is based on the thoughts of Michael Polanyi
[Polanyi, 58] which defined the idea of tacit knowledge
as personal knowledge bound to humans. This type consists of mental models,
beliefs and perspectives [Nonaka, 95]. It is partially
unconscious and therefore difficult to be communicated and explained by
the persons who possess it.
Explicit knowledge on the other hand is formal, codified, systematic,
articulated in writing/numbers, easy to communicate, and shared [Hopfenbeck,
01]. This also means that it can be transmitted and stored for reuse
by other people. Books, documents, data bases and graphs are just a few
examples of this knowledge type.
4.2 Knowledge Conversion
From the business process perspective, the conversion of knowledge into
other knowledge types plays a major role. The conversion between knowledge
types is performed through interaction of tacit and explicit knowledge.
In their book Nonaka and Takeuchi identified four types of knowledge conversion
(see figure 2).
Internalization is the conversion of explicit knowledge into tacit knowledge.
It is very closely related to learning-by-doing. Experiences made through
socialization, externalization or combination are internalized and integrated
into one's own knowledge framework. By this, they can become know-how or
mental models and according to this, very important knowledge assets.
Externalization is the conversion from tacit to explicit knowledge.
By using metaphors, analogies or models one can express his tacit knowledge
in a manner which can be understood by others. It is the essence of tacit
knowledge which can then be handed over in a written form, yet it can be
very difficult to externalize tacit knowledge, often it is simply impossible.
Socialization is a conversion from tacit knowledge of one person to
tacit knowledge of a different person. Often it is done by sharing experience:
Just like apprentices of a craftsman learn their skills by observation,
a knowledge-worker can learn his needed abilities through on-the-job training.
The socialization does not even require speaking or writing a single word.
Figure 2: Model of the Dynamics of Knowledge Creation
Combination is the conversion from explicit to explicit knowledge. Different
kinds of explicit knowledge can be combined through media like telephone,
mail, word processing, further by reconfiguring, categorizing and adding
new information and context to the knowledge.
According to the authors, the creation of organizational knowledge needs
all types of knowledge conversion. To stimulate the process of knowledge
creation in a company knowledge management plays an important role.
The model proposed by Nonaka and Takeuchi establishes a logical framework
which can be used to take a look at tacit and explicit knowledge, the conversions
between those kinds of knowledge and therefore the creation of knowledge
and the conditions and requirements for conversion to happen. It will serve
as the basic framework for modelling a dynamic process of knowledge creation
within the authors' approach.
4.3 Critical evaluation of existing Knowledge Modelling Languages
Knowledge-intensive business processes are characterized by activities
which change, which contain knowledge demands, which can not be planned
easily, and which contain alternative results. Conventional process modelling
approaches do not fulfil all requirements that have to be considered for
modelling knowledge-intensive business processes [Remus,
Gronau [Gronau, 03] proposes a list of requirements
that have to be fulfilled for modelling these knowledge-intensive business
- Goal: Which goal is to be achieved by the model? Are there only documentation
purposes or are a weak spot analysis and the definition of a new process
- Integration of process and knowledge modelling: There should be a
unique approach that combines or integrates the process definition with
the flow and transfer of knowledge.
- Tacit knowledge: Which definition and appreciation of knowledge is
used by the model's approach? Is there a differentiation between explicit
and tacit knowledge? Is it possible to express different levels of tacit
knowledge [Snowden, 00]?
- Knowledge conversion: Are different mechanisms of knowledge conversion
considered and expressed separately in the process model?
- Knowledge flow: Is there a differentiation between information flow
and knowledge transfer?
- Offer and demand: Is it possible to show differences in the model
between the supply of knowledge and its demand?
- Person-related knowledge: Is the modelling of knowledge restricted
to organizational units or is it possible to show knowledge bound to persons?
- Comparison of intended and actual level of knowledge: Is it possible
to compare the knowledge levels required for posts with the knowledge persons
- View representation: Is it possible to navigate through the models
using different views, e.g. an organizational or a process flow view?
- Knowledge maps: Is it possible to generate knowledge maps from the
results of modelling?
Based on these requirements common process modelling approaches like
ARIS [Allweyer, 98], [Scheer,
98], Income [Remus, 02a] and PROMOTE [Karagiannis,
02] were evaluated.
The result of the analysis shows that all analyzed process modelling
approaches do not separate tacit knowledge from explicit knowledge and
that there are deficits in the conversion of the knowledge types and the
person-related knowledge modelling of knowledge in the evaluated approaches.
Major disadvantages of two of the approaches described by can be illustrated
by the following examples. In the ARIS approach the source of knowledge
can not be related to the knowledge and therefore a statement about the
interaction between tacit knowledge and explicit knowledge is not possible.
The Income Process Designer does not support the modelling of knowledge
flow and knowledge conversion.
The result of this evaluation leads to the formulation of requirements
for the specification of a new description language [Gronau,
5 KMDL Object Model
This section introduces the defined objects and their attributes of
KMDL in the actual version 1.1. Furthermore a practical example for modelling
with KMDL is described to support a better understanding of the defined
5.1 Knowledge and Information
Within KMDL the term knowledge is conceived as bound to persons. This
kind of knowledge - tacit knowledge (see section 4.1) - is personal and
cannot be transferred to a formal notation. It is anchored in the activities
and skills of the knowledge carrier and additionally in her/his ideals,
values and experiences [Nonaka, 95]. In contrary explicit
knowledge is easy to formalize.
To realize a clear distinction between tacit and explicit knowledge,
the KMDL differentiates between knowledge and information objects. In KMDL
the term knowledge object refers to the tacit knowledge and the term information
refers to explicit knowledge. New knowledge and information objects are
generated by converting existing elements within the process. This conversion
is based on the interaction between knowledge and information objects.
It has to be noted, that knowledge objects always refer to persons. In
analogy to Nonaka/Takeuchi [Nonaka, 95] the KMDL distinguishes
four kinds of knowledge conversions.
5.2 The Objects of KMDL V1.1
The KMDL provides an object library containing the basic objects "Information
Object", "Task", "Role", "Task Requirements",
"Person", "Knowledge Object", and "Knowledge Descriptor"
[Gronau, 04a]. The connections of these objects are
realised by using a directed information flow as an edge and the four kinds
of knowledge conversion, as introduced in section 4.2. For all of these
objects the attributes identifier, description, keywords, process description
exist. Furthermore for each of the objects optional attributes are defined.
Figure 3 shows the objects and their relations.
The capturing of processes is supported by the definition of aggregated
objects. The process modeller has the possibility to define parts of the
knowledge-intensive business process if required or not. The following
aggregated objects are available: group, role aggregation and task aggregation
as well as process interface.
Figure 3: Objects of the KMDL Data Model and their Relations
The information object is next to the existing knowledge object the
base for the creation of new knowledge objects. Information can be externalised
in an easy manner. It is stored on electronic media or written down in
documents. The creation of new information is done by externalisation or
combination. One characteristic of knowledge-intensive business processes
is the processing of information.
Within the KMDL the input and the output of tasks are represented by
information objects. The specified optional attributes are location, medium,
expiration date, and level (state of working progress).
Tasks are the basic framework for business process models. The sequence
of the tasks determines the temporal structure of the process. A task is
defined as an atomic transfer from input to output, represented as information
Tasks are related to and are fulfilled by (job) positions. The organisational
view of the information flow within the KMDL is workflow-oriented. Because
of that the role is allocated to the task despite of the position.
Roles are taken by persons and have the knowledge objects of all persons
assigned to them. By relating employees and tasks to a position, the functional
and organisational structure of a company can be represented. The optional
attributes are personal data and position. Persons are the owners of knowledge
objects that are necessary to fulfil tasks. The knowledge objects of a
person with the respective knowledge level should be equal to the requirements
of the task the person has to execute.
Performing tasks describes requirements on the roles that are modelled
as task requirements. The totality of task requirements defines the tacit
knowledge that is necessary for a position working on a concrete task.
More than one task requirement can be associated to a role, because normally
more than one capability is necessary to accomplish the task. Here the
knowledge descriptor and level (described within a competency matrix) are
the available optional attributes.
A knowledge descriptor describes the borders and contents of a knowledge
domain and defines partial domains if necessary. It is not codified knowledge.
Task requirements and knowledge objects refer to a certain knowledge descriptor.
The attributes of the task requirement and the knowledge object contain
the required knowledge level within the considered domain. Because of the
definition of the knowledge description the comparison of the desired task
requirement with the available knowledge object is possible.
A knowledge object describes the knowledge of persons. Each knowledge
object must have a reference to a knowledge descriptor for describing which
part of a knowledge domain is covered in which quality. Every used and
needed tacit capability is represented by a knowledge object. In the KMDL
specification, the optional attributes for knowledge objects are knowledge
descriptor, knowledge level (described within a competency matrix), frequency
of access and topicality.
Additionally the different opportunities of knowledge conversion can
be modelled with KMDL, so that the flow of knowledge between persons can
be visualized. Knowledge flows in a process and the different kinds of
knowledge conversion can be used in the model to retrieve information about
the generation of new knowledge and possible weak spots.
KMDL also offers extended representation possibilities to grasp further
characteristics [Gronau, 04b]. These can be transferred
to other expressions of knowledge conversion.
- Frequency: The contact between two persons for the exchange of knowledge
is possible once, often or permanent. The last possibility occurs especially
during an imitation. The other cases can be explained with single or multiple
- Completeness: The completeness of the socialized knowledge has to
be considered. Different or supplementary contents can be given in different
contacts. In addition a complete transfer of the actual knowledge is possible
in every contact.
- Number of participants: A conversion can take place with multiple
participants. A talk given to three people is a single act of socialization.
If this is modelled as three different relations between speaker and listener,
it is meant that three different contacts with three different acts of
- Direction of conversion: A discussion, a brainstorming meeting or
a personal suggestion of one of the participants implicates a multitude
of knowledge flows. These are not directed. Every participant can be either
sender or receiver. Otherwise the acts of socialization had to be represented
on the level of single sentences. Such a degree of detail is not efficient
and no real gain of information. Therefore a representation of expressions
of knowledge flows is necessary, where the participants can be sender,
receiver or both.
The conversion is represented as a node, with that all participants
(knowledge or information objects) are linked. These relations are directed
and show the status of the element as sender or receiver. The line style
shows the frequency of participation while the completeness of the conversion
is represented by the shape of the node symbol. The following figure 4
shows the defined edges and their properties.
Figure 4: Representation of Knowledge Conversion and Objects
5.3 Practical Example
This paragraph describes the modelling of knowledge-intensive business
processes with KMDL using a real world example. It consists of capturing
processes in an international operating (small and medium sized enterprise)
software company. The company's software development is based on standard
products which can be adapted to customer requirements.
While using the software or when introducing a new software component
the customer sometimes recognizes new requirements accordingly to the product.
The company verifies these requirements. One possibility would be to realize
these requirements as a customer specific feature of the software system.
This is usually done when the new feature is specific to the customer's
demands. When the company realizes that the solution is not customer specific
and is demanded by several companies the requirements could be realized
as a new feature in a new release of the software system or as an add on
for the existing one. It could also be possible that the company is not
interested in supporting the customer and does not realize the required
The process described below is part of the software development process
that is carried out when integrating new features in the standard software
product of the company (see figure 5). The whole process was acquired in
the software company mentioned above.
Figure 5: Development of a new Feature in a Standard Software
The integration of new functionality into the standard software product
requires several activities. The identified activities here are functional
definition for the realization, the actual realization of the new functionality
and the test of the new implemented functionality and its integration in
the software product. The "realization of functionality" is represented
as task aggregation which simplifies the modelled process and focuses on
the intended one. All other tasks are specified in more detail. As described
above we just focus on a part of the whole process. This also means that
tasks like "acquisition of customer requirements" are not included
in this part of the model.
In order to carry out the task "definition of realization"
the information objects "functional requirements", "schedule",
"usability requirements" and "integration requirements"
serve as input for the task and the information object "requirements
specification" is the resulting output information object which should
specify all requirements in order to realize the new feature.
The two roles "Technical Consultant" and "Java developer"
are required in order to carry out the task. The requirements for the role
of the "Technical Consultant" are "customer requirements",
"product knowledge", "English competence" and knowledge
about the "system interfaces". As can be seen in the figure the
person "TC_A" in the role of a "Technical Consultant"
has knowledge about the customer requirements, knowledge about the software
product itself and English competence. The person "TC_B" in the
role of a "Technical Consultant" on the other hand has knowledge
about the software product and knowledge about the system interfaces. Together
the two persons satisfy the requirements of the task but both of them are
required because each by oneself does not have the complete required knowledge.
It can also be seen that the person "TC_A" has a monopoly in
the knowledge of the customer requirements.
The second role required in order to carry out the task is the "Java
developer". The role requires knowledge in Java development and about
the system interfaces. As can be seen the person "D_A" in the
role as "Java developer" satisfies the requirement "Java
development". The required knowledge object "system interface"
has to be gained by socializing the knowledge from the person "TC_B"
who has the required knowledge.
There are two additional knowledge conversions identified and displayed
in the modelled process. Through internalization of the project's schedule
the person "TC_A" gains knowledge about the progress of the project.
The person "D_A" defines or refines the coding guide which is
an externalization of the knowledge about Java development.
6 Procedural Model
A detailed capturing and analysis of knowledge-intensive business processes
are required to determine the potentials for improvement in the process.
The Procedural Model ensures the correct elicitation of all data and information
needed (see figure 6). The model consists of six phases. First of all,
it is necessary to identify the knowledge-intensive processes. For this
selection, a criteria catalogue can be utilized. It consists of up to thirty
properties of knowledge-intensive business processes to support their definition.
In the next phase, the capturing of the knowledge-intensive business process
is executed. Here, the model offers a sub-procedure, which contains the
six steps: definition of tasks associated to the process, identification
of the information in- and output, assignment of the persons to the specific
roles, executing the task, specification of the role requirements, and
assignment of the knowledge objects to the accompanying person. The third
phase models the process using the tool K-Modeler. Its practical application
and benefits will be discussed later on. The results of the previous phase
are required for the generation of a qualified concept, which could for
example contain process improvements. The last phase of the model is the
implementation phase, which is only used when Information Technologies
are getting implemented. Participation is an inseparable element of the
KMDL Procedural Model. During each phase the contribution of the participants
Figure 6: KMDL Procedural Model
A detailed capturing of the knowledge-intensive business process is
a pre-condition for the analysis and evaluation of potentials within the
process. The analysis of the process comprises the identification of knowledge-intensity,
the process schemes and the process potential (weak spots).
As-is models illustrate the ownership, the demand, the development and
the use of knowledge. Therefore it is possible to visualise the knowledge
intensity as a kind of knowledge map of the whole process, of a process
part or of single activities of the respective tasks. This procedure enables
the classification of single tasks or the weighting of their relevance.
The results are used for recommendations of technical and organisational
The comparison of as-is models of different instances from the same
process is useful in order to generate universally valid sentences about
process elements and element relations. Special knowledge based activities
should be investigated to identify specific patterns. It is recommended
to extend the existing reference processes with this information in order
to support future participants of the process.
Based on the KMDL approach described above, the KMDL modelling tool
K-Modeler is under development. It allows to model knowledge-intensive
business processes in an easy and intuitive manner as defined in KMDL.
The K-Modeler also supports mechanisms to analyze the processes and generate
reports from the model.
The K-Modeler is engineered using the graphical integration platform
Eclipse [Eclipse, 04]. Eclipse has been developed
to build integrated development environments (IDEs) and already comes with
a variety of core services in order to easily integrate own IDEs with slight
7.1 Integration into Eclipse
As mentioned above the K-Modeler will be integrated into Eclipse. Functionality
is contributed to Eclipse in form of pluggable components, so called Eclipse
plug-ins. The architecture and integration of the K-Modeler can be seen
in figure 7.
Broadly seen the architecture can be divided in three parts, the graphical
layer, the application layer and the persistence layer. The graphical layer
includes all components that are visible on screen like an editor for editing
the KMDL models, views to display properties and attributes and other aspects
of the model and its objects. The application layer provides functionality
for analyzing the model, for syntax checking, for report generation and
other functionality that is processed in the background and not directly
visible to the analyst. The persistence layer provides functionality to
store the model persistent. The model is stored in a relational database
management system (RDBMS). However the persistence layer will be implemented
independently from the storage system which keeps a large degree of freedom
in the choice of the storage system, which means that the RDBMS can be
easily replaced by storage into XML files.
Figure 7: Integration Architecture of the K-Modeler in Eclipse
7.2 Functional Overview of the K-Modeler
In following section the basic functionalities of the K-Modeler is described.
The first section explains of the modelling processes with the K-Modeler
that is carried out. The next part deals with the process analysis that
can be done with the K-Modeler. Then the actual supported process views
are introduced. Finally the opportunity of the skill management support
and the model reuse are explained.
7.2.1 Modelling the Process
The modelling with the K-Modeler is done directly via drag and drop
on a graphical user interface (see figure 8). The analyst can easily model
the process by using predefined graphical elements representing the objects
defined in KMDL like tasks, information objects, knowledge objects, roles,
persons, or requirements (see figure 4). Attributes can be associated with
each object. It is possible to modify and extend the list of attributes
by self-defined attributes. The previously defined types of knowledge conversions
(externalization, internalization, socialization, combination) and the
information flow can be modelled with different types of pointed connections.
The connections between objects without direct interaction are undirected
connections (e.g. the connection between task and role).
7.2.2 Process Analysis
A syntax check for the model ensures that only formally correct models
can be modelled by the analyst. This is done by predefined syntactical
rules. The K-Modeler automatically identifies and evaluates various design
patterns in the modelled processes and thus helps to analyze the process.
These evaluation patterns are derived from known disadvantageous process
elements and structures found in knowledge-intensive processes [Brown,
98]. When applying these patterns in an analysis, the following process
potentials can be identified:
- knowledge monopolies: a single person owns knowledge demanded in the
- unsuitable knowledge profile: no person is modelled in the process
that has the demanded knowledge
- creation of unused knowledge: the created knowledge is not demanded
in the process
Figure 8: Planned K-Modeler integrated in Eclipse
The places of knowledge creation and conversion as well as knowledge
processes like knowledge distribution, knowledge creation and knowledge
use are visible in the model.
With the report functionality of the K-Modeler the analyst can focus
on individual aspects of the modelled process and get statistical results
about the model. The reports can be archived in HTML (Hypertext Markup
Language). Examples for planned reports are:
- what knowledge has been expatiated by knowledge conversion
- the tacit knowledge identified in the process
- results of the analysis of the potentials identified in the process
7.2.3 Different Views on the Model
The K-Modeler provides different views on the process in predefined
abstraction levels. This allows dissolving the aggregations and focusing
on their components.
The following views on the model are distinguished:
- The task view displays the base structure of the process focussing
only on the tasks in the process. A clear distinction between the tasks
in the process is important in order to relate the roles to the requirements
as well as the knowledge objects to the persons that hold these roles.
- The simple process view extends the task view by the information objects
needed to process the tasks.
- In the extended process view the roles which execute the tasks are
also displayed. With this additional information the analyst can identify
the roles that are assigned to the tasks and can identify incorrect assignments
or unintended multiple assignments.
- The tacit knowledge view displays all roles with the assigned actors
and their knowledge objects and requirements that are modelled in the process.
- The general view shows all objects modelled for the process.
The differentiation made in the views is the representation of the object
types and therefore the aggregation.
In addition the K-Modeler allows to group tasks, which enables the analyst
to create task aggregations and dissolve them in order to decrease the
complexity of the view on the process.
7.2.4 Support for Skill Management
The tacit knowledge perspective on the model contains the information
about a persons' knowledge objects and the requirements of the roles. It
can be used for skill management. This allows examining gaps between the
requirements and actual existing tacit knowledge. This information can
be used to plan training processes within the company in order to enhance
the skills of these employees. When a person often demands knowledge objects
of a special topic in a process, an analysis of the situation can be used
to define a conception for further vocational training.
The data tracked via the process modelling can also be used to create
knowledge maps (topic taxonomy) or yellow pages for the company. This in
turn enables the company to identify core competences and experts.
7.2.5 Reuse of the Model
The K-Modeler allows exporting the modelled process into structured
XML. This allows further processing of the gathered information about the
process and can be used for documentation purpose in addition to the reports
or to create yellow pages.
The current approach of capturing processes within the actual practical
projects employs Microsoft Visio. Because of the XML data description of
the objects model it is possible to import these models. Therefore the
further use especially for process evaluation is guaranteed.
8 Practical Experiences
There are several practical projects in which the KMDL was used for
modelling, analysing and improving knowledge-intensive business processes.
The following section gives a short introduction in the objectives and
results of these projects.
In the first application, a big German component supplier uses the KMDL
to capture its quality management processes at the reference processes
specific level to the organisation. The objective was a better integration
of existing knowledge management applications within the company. The component
supplier plans the improvement of access to quality management processes
via KMDL models and other appropriate tools.
With the assistance of KMDL, a German producer of groceries investigates
its information and communication relationships between the customer care
department and the product development. By the KMDL analysis of the captured
process it could be observed, that there was no formalised connection of
information between the question and topics which occurred in the customer
care and the knowledge of existing new products in the product development.
The result of the analysis was the development of a concept for implementing
an Intranet-based tool.
One of the practical projects was in the area of E-Government. In the
context of introducing an intranet in the county, selected knowledge-intensive
business processes were modelled and analysed. By doing this, it was possible
to design the required knowledge management functionalities. Furthermore
the modeller was able to identify technical and organisational process
improvements. The results of the KMDL analysis were part of the conceptual
and technical configuration of the Intranet.
Currently the research group of operational knowledge management uses
the adopted KMDL specification V1.1 for capturing knowledge intensive business
processes. Because of the experiences in the mentioned practical projects
and further research a new version, the KMDL 2.0 is under development.
The present work concentrates on the realisation of a model-driven procedural
method to improve the capturing and analysing of knowledge-intensive business
processes. First of all it is necessary to examine existing meta-models
within the business process modelling and the knowledge modelling.
Appropriate concepts should be identified and compared as well as evaluated
with the meta-model of the KMDL V1.1. The results of the analysis conduce
to the determination and specification of the single process models on
different level of abstraction. This method will be supported by the KMDL
modelling tool K-Modeler.
Furthermore, the offered functionalities of the K-Modeler will be extended.
One of these future functionalities is the ARIS-model import. The use of
existing ARIS models within the company supports the analyst by defining
the simple process view.
It is also required to integrate the so called "person-repository".
This enables the modeller to capture the not attributable objects besides
required information and knowledge objects to fulfil the task. Sometimes
the knowledge which exists beyond the knowledge-intensive business process
is very important for the company.
The description of KMDL-models with Petri-nets supports the simulation
of knowledge-intensive business process. At the moment the simulation is
only used in much formalized business processes. It should be investigated
whether the simulation of this process specification is reasonable or there
are other techniques to simulate the processes, e.g. multi-agent systems.
[Abecker, 02] A. Abecker, K. Hinkelmann, H. Maus,
H-J. Müller: Integrationspotenziale für Geschäftsprozesse
und Wissensmanagement (in German), In: A. Abecker, K. Hinkelmann, H. Maus,
H.J. Müller (eds.): Geschäftsprozessorientiertes Wissensmanagement
(in German), Springer-Verlag, Berlin Heidelberg, New York, 2002, 1-22
[Allweyer, 98] T. Allweyer. Knowledge Process Redesign,
[Brown, 98] William J. Brown, Raphael C. Malveau,
Hays W. McCormick, Thomas J. Mowbray: AntiPatterns: Refactoring Software,
Architectures, and Projects in Crisis, John Wiley & Sons, 1998.
[Davenport, 95] T. Davenport et al.: Improving Knowledge
Work Processes, Working Paper, 1995,http://www.kmadvantage.com/docs/KM/Improving_Knowledge_Work_Processes.pdf
[Davenport, 96] T.H. Davenport, S.L. Jarvenpaa,
M.C. Beers: Improving Knowledge Work Processes, Sloan Management Review,
Summer 1996, 53-65
[Davenport, 00] T.H. Davenport, L. Prusak: Working
Knowledge, Harvard Business School Press, 2000.
[Eclipse, 04] eclipse.org, 2004, www.eclipse.org/
[Eppler, 99] M. Eppler, P. Seifried, A. Röpnack:
Improving knowledge intensive Processes through an Enterprise Knowledge
Medium, In: PRASAD, J. (edt.): Proceedings of the 1999 Conference on Managing
Organizational Knowledge for Strategic Advantage: The Key Role of Information
Technology and Personnel. New Orleans, 1999.
[Goesmann, 02] T. Goesmann: Approach for the support
knowledge intensive processes by workflow management systems (in German),
[Gronau, 03] N. Gronau, U. Palmer, K. Schulte, T.
Winkler: Modellierung von wissensintensiven Geschäftsprozessen mit
der Beschreibungssprache K-Modeler (in German), In: Reimer, U.; Abecker,
A.; Staab, S.; Stumme, G. (edt.): Professionelles Wissensmanagement - Erfahrungen
und Visionen, 2.- 4. April in Luzern, Proceedings der GI, Bonn, 2003, 315-322
[Gronau, 04] N. Gronau; E. Weber: Defining an infrastructure
for knowledge intensive business processes. In: K. Tochtermann, H. Maurer
(edt.): Journal of Universal Computer Science: Proceedings of I-Know '04,
[Gronau, 04a] N. Gronau, E. Weber: Management of
Knowledge Intensive Business Processes. In: J. Desel, B. Pernici, M. Weske,
(edt.): Business Process Management, Springer, Heidelberg, 2004.
[Gronau, 04b] N. Gronau; E. Weber: Modeling of
Knowledge Intensive Business Processes with the Declaration Language KMDL.
In: Mehdi Khosrow-Pour (edt.): Innovations Through Information Technology,
Proceedings of the 14th Information Resources Management Association International
Conference, Idea Group Inc., 2004.
[Gronau, 04c] N. Gronau, C. Müller, M. Uslar:
The KMDL Knowledge Management Approach: Integrating Knowledge Conversions
and Business Process Modeling, In: D. Karagiannis, U. Reimer (eds.): Practical
Aspects of Knowledge Management, Springer-Verlag, Berlin, Heidelberg, 2004,
[Hamel, 90] G. Hamel, C.K. Prahalad: The core competence
of the corporation, Harvard Business Review, 1990.
[Heisig, 02] P. Heisig. GPO-WM - Methode und Werkzeuge
zum geschäftsprozessorientierten Wissensmanagement (in German), In
A. Abecker, K. Hinkelmann, H. Maus, H.J. Müller (eds.): Geschäftsprozessorientiertes
Wissensmanagement (in German), Springer-Verlag, Berlin Heidelberg, New
York, 2002. 47-64
[Heisig, 03] Heisig, P.: Wissensmanagement in industriellen
Geschäftsprozessen (in German). Industrie Management 3, Gito-Verlag,
Berlin, 2003. 22-25
[Hinkelmann, 03] K. Hinkelmann, D. Karagiannis,
R. Telesko: PROMOTE - Methodologie und Werkzeuge für geschäftsprozessorientierten
Wissensmanagement (in German). In: A. Abecker, K. Hinkelmann, H. Maus,
H.J. Müller (eds.): Geschäftsprozessorientiertes Wissensmanagement:
Effektive Wissensnutzung bei der Planung und Umsetzung von Geschäftsprozessen
(in German). Springer-Verlag, Berlin Heidelberg New York Tokio, 2002, 65-90
[Hoffmann, 02] M. Hoffmann: Analyse und Unterstützung
von Wissensprozessen als Voraussetzung für erfolgreiches Wissensmanagement
(in German), In: A. Abecker (edt.): Geschäftsprozessorientiertes Wissensmanagement
(in German), Springer-Verlag, Berlin Heidelberg New York, 2002.
[Hopfenbeck, 01] W. Hopfenbeck, M. Müller,
T. Peisl: Wissensbasiertes Management: Ansätze und Strategien zur
Unternehmensführung in der Internet-Ökonomie (in German), Verlag
Moderne Industrie, 2001.
[Karagiannis, 02] D. Karagiannis and R. Woitsch:
The PROMOTE approach: Modelling Knowledge Management Processes to describe
Knowledge Management Systems. In N. Gronau (edt.): Knowledge management
strategies, processes, communities, Aachen, 2002.
[Kidd, 94] A. Kidd: The marks are on the knowledge
worker, Human Factors in Computing Systems, Proceedings of the SIGCHI conference
on Human factors in computing systems, 1994, 186-191
[Mentzas, 03] G. Mentzas, D. Apostolou, R. Young:
Knowledge Asset Management. Springer-Verlag, Berlin, Heidelberg, New York,
[Nonaka, 95] I. Nonaka, H. Takeuchi: The Knowledge-Creating
Company. How Japanese Companies Create the Dynamics of Innovation, Oxford
University Press, New York, 1995.
[Polanyi, 58] M. Polanyi: Personal Knowledge - Towards
a Post-Critical Philosophy, The University of Chicago Press, Chicago, 1958.
[Remus, 02a] U. Remus. Integrierte Prozess- und
Kommunikationsmodellierung zur Verbesserung von wissensintensiven Geschäftsprozessen
(in German), In A. Abecker (edt): Geschäftsprozessorientiertes Wissensmanagement
(in German), Springer-Verlag, Berlin, Heidelberg, New York, 2002.
[Remus, 02b] U. Remus. Process oriented knowledge
management, Concepts and modeling. PhD thesis, University of Regensburg,
Germany, Regensburg, 2002.
[Scheer, 98] A.-W. Scheer: ARIS From business
process to the application system, Berlin Heidelberg, 3rd edition, 1998.
[Schreiber, 00] G. Schreiber, H. Akkermans, A.
Anjewierden, R. de Hoog, N. Shadbold, W. van der Velde, B. Wielinda: Knowledge
Engineering and Management, The CommonKADS Methodology. The MIT Press,
[Schwarz, 01] S. Schwarz, A. Abecker, H. Maus,
M. Sintek: Anforderungen an die Workflow-Unterstützung für wissensintensive
Geschäftsprozesse (in German). In: Müller, H.-J., Abecker A.,
Maus H., Hinkelmann K. (edt.): Proceedings des Workshops Geschäftsprozessorientiertes
Wissensmanagement - Von der Strategie zum Content auf der WM 2001, Baden-Baden,
[Snowden, 00] D. Snowden: The ASHEN Model an enabler
of action in Knowledge Management, 3(7), April 2000, 14-17
[Trier, 04] M. Trier, C. Müller: Towards
a Systematic Approach for Capturing Knowledge-Intensive Business Processes.
In: D. Karagiannis, U. Reimer (eds.): Practical Aspects of Knowledge Management,
Springer-Verlag, Berlin, Heidelberg, 2004, 239-250