Information Technology Reference
In-Depth Information
over a set of documents, and thus inform the moderator about similar themes that are
discussed in different threads. Standard clustering procedures as well the hierarchical
analysis of textual similarities (Mehler 2002) can enhance the presentation of textual
data in order to support the moderator in formalizing discussion contributions as
reusable experiences or cases.
5
Outlook
indiGo was designed to support all kinds of knowledge that have been identified as
being import for process learning, namely process models (with their associated
templates), experiences from instantiating process models in concrete projects,
discussions about processes in closed or open groups, and private annotations of
process models. Thus with indiGo, any concerned organization member can make
private annotations for a newly introduced, or changed, business process model. Staff
can decide which of the issues that attracted their attention should be discussed within
a selected group of people. The indiGo technical infrastructure enables the
organization of various of such discussion groups based on a customizable discourse
grammar, and indiGo's e-moderation method guarantees that such discussions are
carried in a structured and goal-oriented manner. This helps to identify valuable
experiences, which then are represented as semi-formal cases, and stored in the
experience base. Using case-based reasoning, these experiences are then available for
both process improvement/change and process execution.
The first version of indiGo was presented in March 2002 at CeBIT. Starting in
April 2002, indiGo will be validated within a case study carried out at Fraunhofer
IESE in Kaiserslautern, Germany. New project and strategy processes will be
introduced for the whole institute and indiGo has been chosen as the process learning
platform. We expect very valuable feedback for all the described indiGo methods and
technologies.
In parallel, specified but not yet implemented features will be realized. For
instance, if a process model is modified or reorganized, the corresponding annotations
and discussions should automatically be marked for re-validation or be reorganized
accordingly. In parallel, the indiGo platform will be extended to include the
components on the lower level in Figure 3, starting with CoIN-EF.
As soon as discussions will become available from the case study, text mining
experiments can begin. For that purpose, the discussions in Zeno will be exported in
GXL, an XML dialect for graph structures. Private annotations remain private and
will not be subject to text mining.
Beyond the current project we consider the possibility to extend the indiGo
approach to applications where process models do not play such a central role.
Although a platform for organizational learning should eventually cover all
knowledge categories treated in indiGo, the first steps to organizational learning need
not necessarily involve process models. Maybe, an organization would first like to
invest into an experience base or into a communication platform, and add process
models only later. The challenging research question here is, to which degree
indiGo's methods and technologies can still be applied or easily tailored to such an
organization's needs.
